GENE FLOW IN ANIMAL GENETIC RESOURCES. A STUDY ON STATUS, IMPACT AND TRENDS

Editors: Anne Valle Zárate, Katinka Musavaya and Cornelia Schäfer

Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, Germany

Editors A. Valle Zárate, K. Musavaya, C. Schäfer Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart, Germany

Authors Global study: A. Valle Zárate, C. Schäfer, K. Musavaya, M. Mergenthaler, R. Roessler, H. Momm, C. Gall Global gene flow of : C. Schäfer, A. Valle Zárate Global gene flow of : E. Alandia Robles, C. Gall, A. Valle Zárate Global gene flow of : M. Mergenthaler, H. Momm, A. Valle Zárate Global gene flow of : K. Musavaya, M. Mergenthaler, A. Valle Zárate Case studies: The worldwide gene flow of the improved Awassi and Assaf of sheep from Israel: T. Rummel, A. Valle Zárate and E. Gootwine History and worldwide development of Anglo Nubian goats and their impacts in smallholder in Bolivia: A. Stemmer, C. Gall, A. Valle Zárate Boran and Tuli cattle breeds - origin, worldwide transfer, utilisation and the issue of access and benefit sharing: S. Homann, J.H. Maritz, C.G. Hülsebusch, K. Meyn, A. Valle Zárate Impact of the use of exotic compared to local breeds on socio-economic development and biodiversity in Vietnam: L.T.T. Huyen, R. Roessler, U. Lemke, A. Valle Zárate

Language M. Hill

TABLE OF CONTENTS I

TABLE OF CONTENTS

List of tables III

List of figures III

List of tables, figures and boxes in background information III

Abbreviations VI

Executive Summary 1

1 Introduction 15 1.1 Objectives 15 1.2 Material and Methods 15 1.3 Scope 16 1.3.1 Global study 16 1.3.2 Case studies 17

2 General features of gene flow 18 2.1 Gene flow during domestication and formation 18 2.2 Influence of breeding methods in further development and spread of breeds 20 2.3 Influence of technology and global mobility on the dissemination of genetic material 22 2.4 Gene flow indicated by Country Reports 23

3 Main findings of global study 25 3.1 Global gene flow of sheep 25 3.2 Global gene flow of goats 31 3.3 Global gene flow of cattle 35 3.4 Global gene flow of pigs 42

4 Main findings of case studies 49 4.1 The worldwide gene flow of the improved Awassi and Assaf breeds of sheep from Israel 49 4.2 History and worldwide development of Anglo Nubian goats and their impacts in smallholder farms in Bolivia 53 4.3 Boran and Tuli cattle breeds - origin, worldwide transfer, utilisation and the issue of Access and Benefit Sharing 56 4.4 Impact of the use of exotic compared to local pig breeds on socio-economic development and biodiversity in Vietnam 60 II TABLE OF CONTENTS

5 Analysis of driving factors and impacts 63 5.1 Main factors affecting gene flow 63 5.1.1 Breeding and trade organisations 63 5.1.2 Regulations 68 5.1.3 Foreign aid and development projects 71 5.1.4 Research and commercial interests in specific genes 77 5.1.5 Suitability of breeds for prevailing production systems 79 5.2 Impact of gene flow on economic development and poverty reduction 87 5.3 Impact of gene flow on biodiversity 89 5.4 Limitations of the study 91 5.5 Need for further action 93

6 References 95

7 Contact Addresses 105 7.1 Authors 105 7.2 Advisory panel 106 7.3 Resource persons 107

8 Acknowledgements 112

9 Annex 115 9.1 Background information 115 9.2 Global gene flow of sheep 191 9.3 Global gene flow of goats 229 9.4 Global gene flow of cattle 241 9.5 Global gene flow of pigs 281 9.6 The worldwide gene flow of the improved Awassi and Assaf breeds of sheep from Israel 305 9.7 History and worldwide development of Anglo Nubian goats and their impacts in smallholder farms in Bolivia 359 9.8 Boran and Tuli cattle breeds - origin, worldwide transfer, utilisation and the issue of Access and Benefit Sharing 395 9.9 Impact of the use of exotic compared to local pig breeds on socio-economic development and biodiversity in Vietnam 459

LIST OF TABLES III

LIST OF TABLES

Table 1: Earliest evidence of domestication 19 Table 2: HPI projects with small ruminants in Asia (1985-1997) 73 Table 3: Number of goats distributed by HPI in India (1995-1997) 74 Table 4: Sahiwal crosses with other cattle breeds in Kenya 86

LIST OF FIGURES

Figure 1: Distribution of wild ancestors of major domestic mammals in Western Asia 18 Figure 2: Historical development of breeding methods and impact on gene flow 20 Figure 3: Historical overview on Awassi and Assaf breeding in Israel 50 Figure 4: World wide gene flow of the Improved Awassi and Assaf breeds of sheep from Israel 51 Figure 5: Gene flow of the Anglo Nubian 54 Figure 6: Worldwide transfers of Unimproved and Improved Boran cattle breeding material from Eastern and Southern Africa 57 Figure 7: Worldwide transfers of Tuli cattle breeding material from Southern Africa 57 Figure 8: Performance of Bos indicus, Bos taurus and their crosses in the tropics from two review studies 84

LIST OF TABLES, FIGURES AND BOXES IN BACKGROUND INFORMATION

A 1: Advisory panel 119 A 2: Development of the Merino breed in Europe 119 A 3: Current gene flow in sheep indicated by selected export and import records and records on introduction of foreign genetic material 121 A 4: Composite sheep breeds, number of foundation breeds and the year of origin or year recognised by country of origin 122 A 5: ICAR Report of the working group on recording scheme 128 A 6: Changes in sheep breed composition in the USA 132 A 7: Country Report excerpts for sheep 133 A 8: Country Report excerpts for goats 136 A 9: Possible introduction routes of cattle from West to South, Southeast, East and Northeast Asia, with centres of domestication 139 A 10: Possible introduction routes of cattle into North, West and northeastern Africa 139 IV LIST OF TABLES, FIGURES AND BOXES IN BACKGROUND INFORMATION

A 11: Possible introduction routes of cattle into the Indian subcontinent and Southeast Asia and probable centres of domestication for Bos (bibos) spp. 140 A 12: Cattle migration after the 16th century 141 A 13: Total Simmental population per country** in the early 1990s 141 A 14: German Simmental export to Turkey 142 A 15: EU heifer export by origin and year 142 A 16: Annual EU heifer export by destination 1993-2003 143 A 17: Annual EU heifer import and export 1998-2002 144 A 18: EU heifer import by destination and year 144 A 19: Annual EU heifer import by origin 1993-2003 145 A 20: German Simmental export by year 145 A 21: Introduction of the Simmental breed to different environments 146 A 22: Crosses of Simmental with local cattle breeds 147 A 23: German Simmental export to Balkan countries 1998-2002 147 A 24: Semen export of industrialised countries by origin in 1991 148 A 25: EU semen export by destination in 2003 (%) 148 A 26: EU semen export by destination in 2003 (number of doses) 149 A 27: Semen import of developed countries by origin in 1991 151 A 28: EU semen import by origin in 2003 152 A 29: Semen imports and exports by breed group and region in 1991 153 A 30: Semen imports to developing countries in 1991 by breed, breed group and destination 153 A 31: Annual semen imports and exports of EU in 2002/2003 154 A 32: German semen import (doses) by year and origin 154 A 33: US Brahman export by destination and year 155 A 34: Brazilian Simmental embryo import by origin 1986-1993 155 A 35: Brazilian Simmental semen import by origin 1972-1993 156 A 36: Brazilian semen doses sold in 2002 by breed 156 A 37: Israeli Holstein export by country and time period 157 A 38: AI coverage by breed and country groups for developed countries in 1991 157 A 39: AI coverage by breed and country groups for developing countries in 1991 158 A 40: Use of AI by breed group and region for developed countries in 1991 158 A 41: Use of AI by breed group and region for developing countries in 1991 159 A 42: Share of breeds used in Botswana from 1987 to 1995 159 A 43: Australian cattle registrations by breed 160 LIST OF TABLES, FIGURES AND BOXES IN BACKGROUND INFORMATION V

A 44: Number of by breeds in South Africa in July 2003 161 A 45: Ratio of utilisation of the Simmental breed in different countries for milk and beef 161 A 46: Examples of absorption of local cattle by imported breeds 162 A 47: Holstein’s and Sahiwal’s contribution to composite breeds 162 A 48: Performance of the Holstein breed in different environments 163 A 49: Country Report excerpts for cattle 164 A 50: Annual EU breeding pig import and export 1992-2003 172 A 51: Imported pig breeds during the 1970s into different countries 173 A 52: State of the pig genetic resource utilisation in Uruguay 174 A 53: Country Report excerpts for pigs 175 A 54: Information from the World Watch List of Domestic Animal Diversity 182

VI ABBREVIATIONS

ABBREVIATIONS

ADT Projekt Specialised consulting enterprise that focuses on planning and implementing international agricultural projects (Germany) AI ASPS Sector Programme Support BC before Christ BMPIB Bone morphogenetic protein IB receptor BMZ German Federal Ministry for Economic Cooperation and Development BSE Bovine Spongiform Encephalopathy CAE Caprine Arthritis Encephalitis COMTRADE Commodity Trade Statistics Database (United Nations) CSIRO Commonwealth Scientific and Industrial Research Organisation DAD-IS Domestic Animal Diversity Information System e.g. for example ET Embryo Transfer EU European Union FAO Food and Agriculture Organisation of the United Nations -Africa Food and Agricultural Research Management Organisation (UK) FMD Foot-and-Mouth disease GDR former German Democratic Republic GmbH Gesellschaft mit beschränkter Haftung (limited liability company) GTZ Gesellschaft für Technische Zusammenarbeit HPI Heifer Project International i.e. that is ICAR International Committee for Animal Recording IFAD International Fund For Agricultural Development ISCCD International Scheme for Coordination of Development N North NGO Non-Governmental Organisation No. number NW Northwest OIE Office International des Epizooties PIC Pig Improvement Company Limited (UK) ABBREVIATIONS VII

PPR Peste des Petits Ruminants PrP Prion Protein Gene Locus S South SE South-east SR-CRSP Small Ruminant-Collaborative Research Support Program (USA) TSE Transmissible Spongiform Encephalopathies UK United Kingdom USA United States of America WTO World Trade Organisation

EXECUTIVE SUMMARY 1

EXECUTIVE SUMMARY

The objective of this gene flow study was to describe the magnitude and direction of movement of genetic material of the four major farm animal species cattle, pigs, goats, and sheep over time. Determining factors for these movements should be identified and impacts on economic development and poverty reduction as well as biodiversity in developing countries should be determined on selected examples. No attempts were made to generalise conclusions from sketchy information, but to describe the state of the art avoiding biases. Needs for further actions were to be derived. Historical accounts of the spread of species and breeds are scarce. There is no overview on the current exchange of genetic resources among most countries and within and among regions. The discussion on advantages and disadvantages of gene flow for different stakeholders is controversial. The issue of benefit sharing raised by some stakeholders additionally feeds the political discussion. Therefore, this study intends to provide scientifically founded information for decision makers in the discussion on global gene flow and its impacts. It is certainly not suitable to conclude this discussion. The objectives were approached by conducting separate global studies and case studies for each species. The global studies attempted to understand the historic development and current status of global gene flow in each species. They considered the influence of domestication, breed formation, human migration, breeding methods, technological developments and global mobility on gene flow. Ancestry and historic gene flow were shown for the main mammalian species. Specific breeds were chosen to highlight typical developments. In cattle, the Holstein and Simmental breed were used as examples for Bos taurus, and the Sahiwal and Brahman breed as examples for Bos indicus. Ovis aries was represented by examples from Merino and Hair sheep breeds, Capra hircus by examples from Boer and Angora goats and Sus scrofa subspecies by examples from a breeding company and occasionally from Large White and Duroc. References to other breeds were only made where appropriate. The current status was indicated by exports and imports of genetic material, and influences of foreign genetic material on existing breeds. Results of both parts were depicted for regional clusters. Where information on representative countries was not available, this was noted along with the available information. The case studies examined specific aspects of gene flow. The Improved Awassi and Assaf sheep breeds from Israel were used to depict the historic development and current status of gene flow of these particular breeds, and to summarise their global impacts. The goat case study depicts the history and global development of the Anglo Nubian breed and focuses in its impact study on smallholder farms in Bolivia. The cattle case study gives the origin, global transfer and uses of the Boran and Tuli breeds, focussing on access and benefit sharing. The pig case study includes the impact of exotic pig breeds in Vietnam on socio- economic development and biodiversity. For the historic development, information came mainly from publications and communication with experts. Data for evaluating the current status were obtained from the Eurostat statistic database, Country Reports on the State of Animal Genetic Resources, 2 EXECUTIVE SUMMARY breeding organisations, regional experts and publications. The case studies particularly depended on information from regional experts and organisations. Historical development The majority of the world population of cattle, pig, sheep and goats are found in developing countries, all of them greatly diverse. The development and diversification started after domestication with the spread of domestic animals during the human migrations of the transition. The influence of migration and colonisation on penetration of the Old and New worlds is depicted in regional clusters. Breeding societies formed in 18th century Britain intensified gene flow of animals and diversified new breeds further. In the second half of the 20th century, better global mobility had increased gene flow and complicated its routes. Freight costs, veterinary restrictions, welfare considerations, and exchange and inflation rates have become the main restrictions of gene flow. On the other hand, developing and using has increased gene flow substantially. Differences in the success biotechnology are responsible for different gene flows within the four species investigated. The highest intensity of gene flow is found in semen exchange. The high degree of commercialisation and international competition in the pig sector causes an intensive exchange of pig breeding stock, while comparatively little gene flow is seen in sheep and particularly goats. Differences in the commonplace use of biotechniques between developing and developed countries reflect differences in the access to international genetic resources. In general, the degree of breeding organisation, market power and ability to meet international veterinary and welfare standards limit the supply of animal genetic resources for international markets. Current status Although the current gene flow of all four species is an intensive global exchange, the level varies greatly between the species depending on developments during the last century and the international interest in the trade of the respective breeding products. As mentioned above, it is greatest in cattle, followed by pigs, sheep and finally goats. However, the main stream of gene flow is from developed countries; only in rarely does gene flow originate from developing countries. Destinations are developing and developed countries likewise as both display a significant demand for improved genetic resources. In general, where gene flow takes place, new breeds contribute to diversity by adding to existing breeds and by creating new breed groups and synthetic breeds. The loss of genetic diversity is reported when gene flow completely replaces local breeds with imported breeds. Impact assessments in global terms were not possible for any of the species due to the limited sources of information. However, in analysing the historic development and the current status of gene flow, main factors affecting it were identified. Breeding and trade organisations Countries of origin of gene flow have generally improved breeds as the result of an advanced breeding organisation at their disposal. The high level of breeding and trade organisation in developed countries gives them a considerable advantage over developing countries in exploiting and spreading their genetic resources. In rarer cases, where unique genetic resources with desired genes or traits exist independent of the level of breeding organisation, gene flow originates also from developing countries. Here, important source EXECUTIVE SUMMARY 3 countries demand control over unique genotypes to ensure their market position and long- term benefits. The commercialisation and concentration process in the pig sector developed strong breeding enterprises in developing countries as well. At the same time, unifying national breeding programmes and companies tends to decrease variability in the genetic resources exchanged globally. With increasing commercialisation, gene flow also becomes less traceable as single company transactions have greater impact. In cattle breeding, differences in the AI coverage between developed and developing countries give great differences in access to exogenous gene flow. In general, access to improved genetic material is not dependant on breeding organisations as both developed and developing countries are in demand for improved breeds. However, the potential impact of gene flow in a country increases with increasing degree of breeding organisation. As there are many other factors affecting gene flow, such as governmental interventions and subsidies, it was not possible to identify a relationship between level of breeding organisation and success of gene flow. Regulations Breeding and trade regulations direct and control gene flow. High hygiene standards increase the opportunity of a country to export and globally market its genetic resources. At the same time, such standards hinder gene flow imports and therefore access to foreign genetic resources. Strict import and export regulations may lead to counterproductive illegal transfers of animals. Additionally, governmental policies are the main influence on breeding activities and gene flow. Regulations in trade and breeding include market-regulating subsidies, and other governmental activities influence gene flow by importing selected breeds as part of national plans or projects. Research and commercial interests in specific genes Interest in specific genes affects gene flow by prompting the flow of these genes from their countries of origin to all areas where a commercial or research interest exists. General interests focus on reducing production costs and improving product quality. Serving consumer preferences plays an important role in this context. Breeds from countries with traditional production systems move into the centre of interest as they harbour alleles lost in countries with intensive production systems. Foreign aid and development projects Foreign aid and development projects like national development projects can start or prevent gene flow depending on whether crossbreeding or replacing indigenous stock is promoted, or genetically improving local breeds. A widespread approach to increase the national output from animal production in developing countries is to introduce improved temperate breeds for pure- and crossbreeding purposes, starting significant gene flow. However, due to complex reasons, such as genotype-environment interactions and local socio-economic conditions, efforts varied in their success. The suitability of the introduced breeds to climate is a prerequisite for the success of a project and the sustainable impact of gene flow. Compared to commercial activities, the impact of gene flow started by development projects has generally been low. New approaches focused on procuring improved breeds of similar climates, or dual-purpose breeds, or of exotic and local breeds. Funding improvements to local breeds is a modern activity resulting from 4 EXECUTIVE SUMMARY failures to establish exotic breeds. However, impact expectations are also limited due to the low amount of funding. Suitability of breeds The suitability of imported breeds for prevailing production systems and environments determines their impact on existing breeds. If they are suitable, gene flow establishes new genotypes in existing livestock populations, either by coexisting with the existing breeds or by replacing by direct interference or market competition. Small and large ruminants depend strongly on their production environment, so the prevailing production conditions largely determine the breeding or production purpose and the suitability of breeds and breeding methods. A second determinant is the climate zone for which the imported breed is meant. There are distinct breeds and breed groups suitable for diverse purposes in the different ecozones. In commercial pig production, the environment is more often shaped according to the needs of the pigs so adaptability problems are often reduced. The worldwide gene flow of the improved Awassi and Assaf breeds of sheep from Israel In Israel, within-breed selection in the unimproved sheep started at the beginning of the 20th century by Jewish sheep breeders, resulting in the formation of the Improved Awassi strain. To improve its prolificacy, the Improved Awassi was crossed in the 1960s with the East Friesian Milk sheep that was imported from Germany, resulting in the formation of the Assaf with improved prolificacy. Today, the Assaf has nearly replaced the Improved Awassi in Israel’s intensive dairy sheep sector. Recently, the Booroola gene, a major gene coding high prolificacy, was introgressed by crossbreeding to the Improved Awassi and the Assaf resulting in two new strains, the Afec Awassi and the Afec Assaf, respectively. The detailed review of gene flow of the Improved Awassi breed from Israel showed that since its start in 1965, 28 transfers to 15 different countries are documented, with summarised monetary transfers over $2 million. Totally, 5,433 lambs, 1,100 doses of frozen semen and 143 embryos of the Improved Awassi have been exported from Israel. While only 9 of these 28 transfers have been part of development aid projects, the majority represent commercial transfers between private sheep or government institutions and Kibutzim in Israel. Improved Awassi breeding material was transferred to Southern and Eastern Europe, Central Asia and Australia. The transfers to the tropical countries Burma, Ethiopia and India were part of development projects. In the Middle East, a total of 1,113 Improved Awassi lambs have been exported to Jordan, Iran, Abu Dhabi and to Turkey, where most of them were used to improve local stocks of unimproved Awassi. Unofficial transfers to secondary destinations could not be traced accurately. The gene flow of the Assaf started in 1977. 687 lambs, 11,354 doses of semen and 260 embryos, for a total price of $333,040, have been exported from Israel in 10 transfers to 7 different countries. Only 2 transfers were part of development aid, while the majority were commercial imports. The main stream of the gene flow of the Assaf breed of sheep went to the Iberian Peninsular. In these countries the Assaf is today the dominant dairy sheep breed, numbering over 1.2 million pure- and crossbreds. Besides those two countries, the Assaf has been transferred to Peru, Jordan and Abu Dhabi, but with much less effect on the sheep production sector. From Portugal, Assaf breeding material was exported to the EXECUTIVE SUMMARY 5

United Kingdom and Italy. No Assaf breeding material has been transferred to Eastern Europe, Central Asia, Australia or New Zealand. History and worldwide development of Anglo Nubian goats and their impacts in smallholder farms in Bolivia The Anglo Nubian is an example of a breed developed by combining genetic resources from different parts of the world joining performance and adaptation to tropical conditions. It is a dual-purpose goat used for milk and meat production. Herdbooks of the Anglo Nubian are kept in Britain, the USA, Canada and Australia. The countries of the South with purebred or crossbred Anglo Nubians have no such official records and information on numbers is scarce or non-existent. The initial crossbreeding, which led to the formation of the Anglo Nubian breed took place during the latter half of the 19th century. In 1910, the Anglo Nubian was recognised as a breed in England and registry began. Anglo Nubians were exported for the first time from Britain to the USA in 1909, reaching a total of about 30 goats up to 1950. Here, Anglo Nubians were bred and selected without any further crossbreeding with other breeds. In Canada, a breeding programme was established in 1921, based on imports of Anglo Nubians from Britain. Offspring were imported into the USA and continued to have a great impact on Anglo Nubians there until the late 1940s. From the USA Anglo Nubians were exported to Puerto Rico and Latin America as early as the 1940s. Later on, Anglo Nubians were exported from Britain and the USA in several development efforts in Latin America, Africa and Asia. In some countries, Anglo Nubians continue to be kept as (Mexico, Brazil, Peru, Colombia, several Caribbean states, Egypt, Israel, Oman, India, Bangladesh, The Philippines, Mauritius and Malaysia) although numbers are sometimes so small that it is difficult to preserve the population (Venezuela, Ecuador, Thailand). More widespread is the use of the Anglo Nubian in crossbreeding. In some countries, like Cuba, imported Anglo Nubians together with other specialised breeds caused a decline in number of the local Criollo goat. In Bolivia, goats are mainly kept in the inter-Andean valleys at altitudes ranging from 1,000 to 3,000 meters altitude. Anglo Nubian goats were imported to Bolivia in the late 1960s up to the present. Animals originated in Argentina, Brazil, Paraguay or the USA; semen was introduced from Germany. There have also been exchanges between different parts of the country. Anglo Nubians were introduced in the regions with a tradition in goat keeping, namely the departments of Cochabamba, Tarija, Chuquisaca, Potosi, and some parts of Santa Cruz. The majority of imports and exchanges within Bolivia were planned and handled by development agencies (6 cases), non-governmental organisations (3 cases) and universities (2 cases) while only in a few cases did individual goat owners imported Anglo Nubians. Introduction was successful in intensive or semi-intensive production systems, whereas no benefits were observed in semi-extensive and extensive production systems. Here, in cases of persistence of the breed, it could often not make use of its production potential and was therefore not superior to Criollo goats. However, introducing Anglo Nubians in intensive production systems had positive side effects through improved pasture and herd management. The impact of the introduction 6 EXECUTIVE SUMMARY and promotion of the Anglo Nubian in Bolivia on the local goat population in predominantly very extensive production conditions was largely unsuccessful and unsustainable. The focus on Anglo Nubians in development projects and research resulted in a waste of resources invested and limited the allocation of resources for Criollo goat improvement but did not cause direct damage. Only in exceptional cases, under more intensive management conditions, positive economic and environmental effects could be stated. Boran and Tuli cattle breeds - origin, worldwide transfer, utilisation and the issue of access and benefit sharing African indigenous cattle breeds, particularly the Boran and the Tuli, have received increasing interest in the past as a source of genetic diversity with potential to improve cattle production in sub-/tropical environments worldwide. In this context, controversy has arisen about conservation-through-utilisation strategies and access and benefit sharing. The Boran (Bos indicus) is a major cattle breed in eastern Africa, originating in the Borana rangelands in southern Ethiopia. The Tuli cattle (Bos taurus) descend from a small nucleus herd of yellow Ngwato cattle in Zimbabwe. Both breeds have evolved under harsh arid and semi-arid range conditions, and out of the local herders’ indigenous breeding and selection strategies. In Kenya, rapid breeding of the Boran cattle was started by British settlers, leading to the formation of the Improved Boran. In Zimbabwe, the Tuli were further developed at government breeding and research stations from the early 20th century. Breeding organisations were founded in the regions of origin, and research and breeding programmes were set up to improve beef production while maintaining adaptability to environmental constraints. Boran cattle showed high fertility and production in low-input environments, while Tuli cattle also showed high fertility and excellent beef quality, but produced comparatively better in high-input environments. Because of their adaptability and productivity in tropical conditions, Boran and Tuli cattle attracted the interest of livestock scientists and the international beef industry. In 1988, in order to add breeding options to composite populations for industrial beef production, Australian researchers - collaborating with a consortium of Australian beef producers - imported the first Boran and Tuli embryos from Zambia and Zimbabwe. In 1991, Boran and Tuli embryos were exported from Australia to research stations in Nebraska and Texas, USA, where the largest germplasm programmes on in the world were undertaken. The breeds also found their way into the Australian, American, and South American beef industries through various other channels. The majority of documented transfers of Boran and Tuli genetic material represent commercial transfers between government research stations and private enterprises and among international business partners. The Tuli appears to have been better accepted among beef producers in Australia and America than the Boran. However, comprehensive data about the current population and the contribution of both breeds in crossbreeding schemes or in the formation of composite breeds are scarce. From the few available figures, the population of these breeds in these countries is small. Given the overall size of the beef industry, with 94.9 million head of beef cattle in the USA and 26.4 million head in Australia in 2004, the likely contribution of Boran and Tuli cattle to the sector appears insignificant. However, the study shows that a demand exists for genetic material with adaptive traits and special traits of beef quality. EXECUTIVE SUMMARY 7

In Africa, Boran and Tuli cattle were used systematically to a considerable extent for commercial ranch development, but not for local or regional livestock improvement schemes. Today, Boran and Tuli cattle are said to face genetic dilution in their areas of origin, although they are still used - particularly the Improved Boran in Kenya - for commercial beef production. Some Non-Governmental Organisations argue that the Boran and Tuli were exploited by networks of institutions, business companies and individuals, without sufficient compensation to the original breeders. They have raised a controversial debate on access and benefit sharing agreements, which is dominated by three main issues: Whether or not a prior informed consent existed before the initial transfers from Africa into Australia; whether or not the call for additional compensatory payments to the original breeders is justified; and how big the impact of Boran and Tuli genetic material on the Australian and North American beef sector actually was. The first two issues are ethical rather than scientific, and have to be politically resolved. The actual contribution of the Boran and Tuli to upgrading the Australian and American beef sector appears to be negligible. It did not nearly meet the high expectations which triggered the operation, and the actual use of Boran and Tuli cattle in the Australian and American beef sector is limited to singular cases. Prior informed consent should be more carefully looked for in future transactions to avoid a posteriori claims. Impact of the use of exotic compared to local pig breeds on socio-economic development and biodiversity in Vietnam This case study focuses on the formation and distribution of the main indigenous pig breeds and crossbreds in Vietnam, the introduction of high performance breeds and their impact on biodiversity, and the suitability of different breeds for different environments. Vietnam owns a wide variety of local pig breeds across different regions of the country. At present, exotic and crossbred pigs dominate, while local pigs make up only 26% of the national pig herd, mostly in uplands, rural and remote areas. The decentralised Vietnamese breeding system, the less developed central coordination and the common use of AI have all supported the spread of exotic pigs in Vietnam, especially at the smallholder level which makes up 80 to 95% of Vietnamese pig production. Smallholder pig production includes different intensity levels. In contrast with large-scale commercial pig production, they can be characterised as low-input systems. Local pigs yield lower reproductive and growth performances. Performance data in literature are rarely comparable, as local breeds were usually investigated in low-input conditions, while exotic pigs or crossbreds are often tested under improved conditions or on station. The influx of exotic breeds had a strong impact on local pig populations. Today, 10 of 14 local pig breeds are in vulnerable or critical state or face , and all of them show declining populations. The main Vietnamese institution conducting conservation programs does this only for a limited number of pig breeds. The long-term sustainability of those programs is questionable. The significant genetic distinctions both between Vietnamese breeds and between Vietnamese and European breeds have been shown. Local breeds are a source of promising 8 EXECUTIVE SUMMARY alleles, which might be significant for future genetic improvement and of unpredictable economic value. Local pig breeds are a significant component of the Vietnamese and worldwide biodiversity, and are still important for resource-poor farmers in Vietnam, who depend on them to ensure their livelihoods. The of high-yielding exotic breeds will increase in intensified production systems. Local breeds will only contribute to worldwide biodiversity if their competitiveness to exotics is proved for production systems under development and/or if favourable adaptation traits are proved and the controlling alleles identified. Investigations are under way to define local pig breeds, characterise them, and compare their performances under standardised conditions. Impact on economic development and poverty reduction From the main findings, the following conclusions are made about the impact of gene flow on economic development and poverty reduction. Developing countries display a demand for new breeds of all four species to increase productivity of farming. With biotechnology and better transport, genetic resources have become readily available and accessible across the globe. However, the quantity of genetic material transferred is not important to the success of gene flow. Instead it is its suitability to the new environments, economies, cultures and societies. If the introduced improved breeds are suitable for the prevailing conditions and production systems, gene flow contributes significantly to the country’s economy and reduces poverty by improving on-farm productivity. This may include increased red meat production as reported for pigs, sheep, goats and beef cattle. The importance of gene flow for economic development is particularly evident where the local breeds have a low production potential and so cannot be quickly improved through within-breed selection. On the other hand, as imported gene flow often involves specialised and concentrated production the poorer rural population may not benefit from it. The greater the success of the introduced breed, the greater the threat of losing biodiversity in the national population. Many marginal areas, such as deserts, scrublands and mountainous areas, can only be exploited by locally adapted breeds. If these breeds die out, it will no longer be possible to use large areas to produce food at least not in extensive production systems, operated by sizeable numbers of poor livestock keepers. Beside negative environmental impact and the national economic loss, the livelihoods of the farmers and pastoralists of these resource-poor areas are directly threatened as they rely on these local genetic resources and do often not have an alternative to make a living. If the new breeds add to the diversity of national populations, gene flow offers the chance to quickly adapt to new production options. Particularly, in rapidly developing parts of the world, gene flow vitally enhances production to meet the population’s growing demand for red meat. Production can be further improved if this development goes hand in hand with a change from extensive to intensive production systems. These intensive production units will compete through the market with extensive production systems, so threatening the livelihoods of smallholders. The negative effects for smallholders are more severe the more the production systems of exotic and local genetic resources differ, and the more concentrated and commercial the production of the exotic genotypes is. The pressure of market competition rises if governments help to establish large-scale livestock production through laws and incentives that exclude smallholders. However, it is a political decision EXECUTIVE SUMMARY 9 how national resources are allocated and whether or not efforts are made to ensure poor farmers participate in the development process on or off their farms. Left to survival strategies, they are better off with their local breeds than without them. To get out of poverty, relying on local breeds will hardly ever suffice. Restricting access to advanced technology, here specifically access to high performing exotic breeds, is likely to hinder self-sustained development. However, protecting the developing local breeding organisations and avoiding hidden subsidies to exotic competitors may help local livestock breeders to participate in economic development. Measures to counteract concentration in the breeding organisations and the trends to monopolise access to genetic resources may also help. The choice of technology, here the access to breeds and genetic material, cannot replace political decisions. Gene flow may also contribute to niche market production, permitting or even encouraging the use of foreign breeds with characters specific to local market preferences. Niche markets provide a promising chance to survive, particularly in which otherwise faces severe reductions or even lost profitability. On the other hand, native genetic resources may contribute as much to niche market production or other production purposes which also conserves national genetic resources. A particular importance of gene flow is given when livestock populations are reduced, for example after wars. When effective breeding work is limited and the basis to restore and develop future populations is lost, gene flow is the only chance to rebuild the stock from imported genetic material. However, a decision has to be made whether the rebuilding should introduce high performance breeds from temperate climates or local breeds from neighbouring regions. If the imported breeds are unsuitable for prevailing production systems, the impact of this gene flow on population composition is theoretically self-limiting. However, the economic losses for smallholders can be severe, if development projects or governments continue to promote or subsidise these genotypes. The losses are caused by diluting or replacing the indigenous genetic resources suitable for the smallholder production system. There is also an indirect loss as genetic resources are wasted that could be used later in suitable genetic improvements. Impact of gene flow on biodiversity The influx of high performing genotypes into existing breeds has always been an important component in developing and improving breeds. In the history of all species investigated, gene flow has contributed significantly to diversity. In population genetics migration is an important source of genetic variability. The growing global need for improved animal productivity has led to an increased demand for improved genotypes. With the biotechnology and global transport developed since the mid 20th century, foreign genetic resources are readily available across the globe. Current gene flow is dominated by source countries with advanced breeding structures, mainly developed countries, so that gene flow is north to north and north to south. At the same time, animals from developed countries increasingly belong to a small number of genetically narrow, high performance breeds with similar breeding goals and hybrids developed over the last two centuries and strongly influenced by controlled research- funded breeding programmes. They have been selected for high yields and require standardised conditions and high inputs to exploit their genetic potential. The international 10 EXECUTIVE SUMMARY exchange of genetic material from a decreasing number of sires increasingly loses genetic variation with global impacts for developed and developing countries. The global gene flow of a relatively few breeds is significant for all species under study. Biodiversity is now recognised as relevant to global politics, and has been too long neglected in research and its economic relevance underestimated. The impact of this gene transfer on biodiversity depends mainly on the breed’s suitability to its new production environments and can be measured in changes to the national population composition. If they are suitable and the new breeds add to existing genotypes, then biodiversity increases. If new genotypes replace existing breeds, biodiversity is lost. Examples for both situations are numerous for all species and in general they are examples of national economic development. Whether the improved breeds co-exist with or replace the native breeds after gene flow depends on various factors. One factor is the production system that the introduced genotype is meant for. In developing countries, where extensive smallholder production systems prevail, dilution or replacement, mainly through crossbreeding, are likely where the new genotypes are suitable, are available, and are adopted by the local breeders. Then, developing smallholder livelihoods can conflict with improving biodiversity and it may not be adequate to require poor farmers in developing countries to bear the costs for conservation of global biodiversity. More often, these introduced genotypes require more intensive production environments and therefore add to the more extensively managed local breeds. However, by competing through the local markets, intensive production units can threaten traditional production systems, and with them the local genetic resources. The more commercial and concentrated the sector is the greater the threat of losing biodiversity, a general rule which can be observed in other sectors likewise. According to data collected by the FAO, 18 % of the 740 farm animal breeds that were recorded as extinct were breeds from the South, although the percentage may be underestimated as data collection started later and is more difficult in developing countries. However, among the breeds at risk, including the status endangered and critical, 60% are from the South and this proportion is expected to increase. The figures differ greatly between sources and between species and their reliability is further weakened by the great number of breeds with unknown status. However, if the risk factors “change of husbandry”, “expansion of large-scale intensive livestock production” or “people giving up herding or farming” are taken into account, then the South could become the hotspot of breed loss of the 21st century. Niche market production using local breeds to serve specific local consumer preferences may counter these losses. Careful production system analysis is therefore demanded to evaluate the impact of gene flow on biodiversity, as well as to determine whether conserving local breeds is in the interest of local farmers. In Europe we can see a re- orientation towards the use of local breeds, due to a new price system based on eco-tourism and conserving natural environments. Niche market production using foreign breeds with specific attributes to serve local market preferences are also examples of the benefits of biodiversity by transferring unique genotypes. The utilisation of major genes is in this context of particular interest. Developing composite sheep populations from a combination of endangered and established breeds may result in financial benefits and simultaneously conserve domestic EXECUTIVE SUMMARY 11 animal diversity of the gene pool, although not of breeds. However, these are singular cases and are not meant to suggest that an overall model solution is to align the global interest to conserve biodiversity with the interests of local farmers to quickly increase production. In general, if global gene flows that reduce biodiversity are to be counteracted, creating incentives for smallholders in developing countries such as paying to conserve global biodiversity through local genetic resources may be more promising than imposing restrictions. If the introduced breeds are not suitable, as mentioned above, the impact on biodiversity is theoretically limited. However, if the introduction is promoted or even subsidised through development projects or governments, or crossbreeding is indiscriminate over a long time and in considerable quantities, existing breeds will be affected. The effect is generally negative as the native and adapted breeds are diluted. However, over a long adaptation time, these new genotypes can also contribute to biodiversity by developing new breeds or traits. In Latin America for example, introducing unsuitable wool sheep breeds from Spain led eventually to the well adapted Chiapas sheep, now considered native. Apart from the main streams of gene flow of high-yielding breeds from developed countries, gene flow takes place from all countries which possess unique genotypes. High- yielding breeds formed in developed countries depended on foreign genetic material. For example the European and North American commercial pig breeds were heavily influenced by genetic material from Asia. Research and commercial interests focus at such genotypes of current and future importance. Global transfer of genotypes generally contributes to biodiversity, if these genotypes are threatened in their native countries. By making use of their typical adaptability to marginal production environments, valuable genetic resources from developed countries can be conserved as well as developing economies in developing countries by genetically improving local breeds. Limitations of the study The most limiting factor of the study was the information available on current gene flow. Countries with highly organised breeding structures were expected to have statistical data bases on breeding stock transfer and changes in national population composition. However, only Europe had available statistics in the form of the Eurostat statistic database and then gene flow records were limited. First of all, the records did not include breed details so that the influence of single breeds could not be interpreted. It was also unclear whether these records reflected true volumes of animal transfer, for example whether transfers for development projects were included. Furthermore, only the exchange of purebreeding stock was recorded. Purebreeding statistics are however not important where hybrids play a vital role, particularly in pig breeding. Additionally, with increasing commercialisation, export and import statistics are less able to depict gene flow and company figures would be more suitable. However international breeding companies tend to restrict access to this information at least to up-to-date figures. In commercial breeding, transfer figures underestimate the gene flow when a few animals are multiplied in national centres, as the case in commercial pig breeding. Apart from cattle, data on the exchange of other breeding products, such as semen and embryos, were even more difficult to obtain and in general lacking. Additionally, there 12 EXECUTIVE SUMMARY may be big gaps between what has been transferred and what has been successfully utilised. Another major methodological restriction is that the availability and quality of information on transfers varies considerably between countries and extrapolating from one well documented situation to cover the many information gaps is not possible. Official national statistics were available via the FAO on the exchange of live animals but these were not exclusively for breeding. The country reports in some cases included necessary information. However, if country reports included data on animal transfers, the usual focus was on imports not exports. Exports could only be traced indirectly as imports into other countries. Therefore, an important source of recorded information on gene flow was generally left out. This was particularly bad for developing countries, where records of animal movements rarely existed and country reports, if they existed, did not contain the desired details. In general country reports differed considerably in quantity and quality and no standard format exists which would allow such information to be compared. Export data from developed countries could have closed that information gap. The latter limitation throws light on the lack of information about gene flow from developing countries, particularly South to South movements which would have been interesting. A major constraint is that the size and number of transfers does not generally predict its genetic or economic impact. Transactions are, with very exceptions, impossible to follow up. There were examples for all species where massive transfers had very little long-term effect, and where the transfer of a few animals had major impacts on global gene flow. These examples however only become evident retrospectively as changes to national population composition, making this indicator of gene flow an important tool in the study. It was used especially in the global study but relied mainly on the Country Reports which, as mentioned above, were limited in quantity and quality. Additionally, a major limitation of the country reports on changes in national population composition was put forward in the pig case study. According to more updated sources it appears that the Vietnamese country report did not realise the full influence of exotic genotypes on local genotypes, and only reported where both were directly interfering. So the impact of the exotic genotypes on the local populations was underestimated. With the increase of global mobility and biotechnology, gene flow has increased and has become more complicated and even more difficult to follow up. This puts severe limitations on the global studies. The Awassi case study showed in an illustrative example how this limitation could be overcome in particular, detailed case studies. But it also showed what effort it takes to collect the necessary information to assemble a nearly complete picture of gene flow for one particular breed from one particular regional source. The study as a whole therefore had to refrain from quantifying global gene flow. It was shown that gene flow is a very complex and dynamic process with many factors varying the possible impacts from case to case. The impact evaluation of global gene flow on economic development and biodiversity as treated in chapter 8.3 and 8.4 therefore make very general and qualitative conclusions and do not measure overall global gene flow. The study analyses selected aspects of gene flow and is meant to provide scientifically-founded information for decision makers in the current discussion on global gene flow and its impacts. It may help to understand the complexity of the situation and may prevent EXECUTIVE SUMMARY 13 decision makers from implementing oversimplified programmes. The study is certainly not suitable to conclude this discussion. Need for further action The discussion on gene flow can only be fruitful if the impact of gene flow can be suitably evaluated. It has been stated above that global evaluation is doubtful and that global interpretations will be possible only if serious status and impact studies in each country add to the global picture. With the limitations of the data bases accessed in this study, further actions are needed to improve the information on current gene flow in a standardised comparable format for each species and country if a global valuation is intended. As improving data quality and quantity in developed as well as developing countries, particularly import and export figures, are not realistic in the near future the focus should be to derive reliable data on national population composition changes. This calls for frequent census studies at breed and genotype level. This approach, which for most countries is challenging enough, would avoid the time consuming tracing of complex gene flow routes and instead starts right at evaluating the genetic impact and can be more easily connected to collateral impacts on biodiversity and economic development for different groups of society. Many weaknesses could be minimised, such as the different values put on national genetic resources by different stakeholders. Case studies then need to provide additional information on the impact evaluation for very specific countries or regions as well as on the organisational structures and mechanisms influencing the impact. Each country, based on its evaluation of the national situation, can then make use of gene flow control mechanisms to improve the situation. One has to be aware that indigenous genetic resources are a source of adaptability to specific changes in the economic and natural environment. Conserving these resources secures the national as well as global adaptability to changing future animal production conditions, and secures the livelihoods of animal holders who for the time being depend on these resources. Future animal production has to cope with developing countries’ biodiversity and reduce poverty. Concepts which serve both by securing or even improving smallholder livelihoods in marginal areas using their indigenous genetic resources should be heavily weighted. These concepts need to identify and evaluate the genetic distance between breeds as well as the genetic potential, performance and other uses and non-use values in different production systems. What is more, biodiversity impacts must be taken into account when conditions of gene flow are evaluated. Exotic breeds are currently introduced for various reasons. If the genetic potential of the local breed is low, economic development calls for introducing suitable higher performing breeds. Suitability requires reasonable selection of imported breeds, comparative performance evaluation of different pure- and crossbred genotypes, and evaluation of natural, economic, social and cultural framework conditions, before selecting a genotype and breeding system. Conserving national genetic resources by crossbreeding can be considered as well as using foreign breeds which are under threat in their country of origin. Such approaches should preferably be carried out on-farm closely involved with the livestock keepers. However, if conserving indigenous genetic resources conflicts with economic development and poverty reduction, conservation cannot be carried out at the expenses of local livestock keepers. Conservation then has to be treated separately but alongside economic development, supported by appropriate funding. 14 EXECUTIVE SUMMARY

Herders and farmers must be closely involved in decision making, and appropriate system analysis must be developed for smallholder production systems in developing countries. Based on its central importance to reduce poverty and conserve biodiversity in marginal areas, community based management of animal genetic resources and indigenous knowledge has to be fostered. Each country has to be aware that unique genotypes are in global demand. Gene flow is too complicated to find routes after the event, in whatever quantity. Gene flow should therefore be considered case by case, and recording practices and access and benefit sharing agreements put in place in advance, rather than retrospectively attempting to measure and exploit the transfers. The apparent lack of prior informed consent is one of the major discussion points in the case of the Boran and Tuli genetic resources transfer, as detailed in the case study. Finally it has to be concluded that without local breeding organisations, actively including smallholders in the breeding work is impossible over the long run. Instead, international breeding companies are ready to do that job. As already observed in cattle and pig breeding, a two-level structure is emerging: modern and traditional. What is more, as the demand for meat and milk in developing countries is expected to double in the next two decades, a new “livestock revolution” is expected. In the long run, this rapid expansion of intensive production may put traditional production under increasing pressure. However, the importance of for the livelihoods of smallholders, particularly in marginal regions, cannot be overestimated. If regional structural changes do not allow these groups to find new incomes, then breeding should be organised and payments to conserve natural environments and farm animal diversity should be considered world wide.

INTRODUCTION 15

1 INTRODUCTION

1.1 Objectives The objective of this study is to describe the magnitude and direction of movement of genetic resources of the four major farm animal species: cattle, pigs, sheep and goats. Determining factors for these movements should be identified, along with their impacts on economic development, poverty reduction and biodiversity in developing countries. Necessary further actions were to be derived where results permitted. Historical accounts of the spread of species and breeds are scarce and subjective and include hardly any quantities. There is no overview on the current exchange of livestock genetic resources between most countries or within and among regions. Such a background document is needed to draft sound policies and sustainable programmes, and to conserve livestock genetic resources for future breeding requirements of nations, regions and the globe. The discussion on advantages and disadvantages of gene flow for different stakeholders is controversial. Some studies argue that gene flow from developed countries has done more harm than good and that a variety of livestock breeds have been heavily reduced (Matthias and Mundy, 2005). In other studies, genetic resource loss in developing countries due to gene flow could not be proved, and gene flow’s importance for economic development is stressed (Gibson and Pullin, 2005). The issue of benefit sharing, raised by stakeholders, feeds the political discussion (Sansthan and Köhler-Rolefson, 2005). Therefore, this study intends to provide scientifically founded information for decision makers in the discussion on global gene flow and its impacts. It is certainly not suitable to conclude this discussion. Direct beneficiaries may be political decision makers, particularly in developing countries, and relevant international bodies and donor institutions. Intermediate beneficiaries may be farmers, pastoralists and breeders, whose interests must be considered in international discussions on regulating access to genetic resources and related benefit sharing. This study will be incorporated into the “The State of the World’s Animal Genetic Resources for Food and Agriculture” of the Food and Agriculture Organisation (FAO) as one of the thematic studies. It will also contribute to the “Global Strategy for the Management of Animal Genetic Resources” by FAO.

1.2 Material and Methods The study has been implemented by the Institute of Animal Production in the Tropics and Subtropics of the University of Hohenheim and commissioned by the Federal Ministry for Economic Co-operation and Development (BMZ) and German Technical Co-operation (GTZ). The Food and Agriculture Organisation (FAO) acted as a support agency. An advisory panel composed of international scientists, representatives of donor and development agencies, the private sector and Non-Governmental Organisations (NGOs) closely accompanied the study (Annex 9.1 A 1). Feedback was provided in meetings, and by reports and e-mail advice throughout the study. National and international statistical data were analysed and literature reviewed. For specific information on breeds, breeding organisations were included. The intermediate analysis results were used to interview national and international experts. Further prepared 16 INTRODUCTION excerpts of gene flow and relevant contents of 114 Country Reports on animal genetic resources were incorporated, supplemented by expert advice when interpreting results (Annex 9.1 A 1). Initial reviews of available information identified information sources, uninformative sources and information gaps. Only data from the Eurostat statistical database of the European Union was used in detail. It showed from the perspective of Europe which countries and regions mainly exchanged breeding animals and when. However, as breed information was not included, and there is little international data on important export nations like USA and Australia, conclusions are only possible for limited areas. Additional breed information was obtained from breeding organisations. Extensive statistics from developing countries were not available. The Country Reports differ strongly in amount and quality of information. Comparable quantities could not be obtained for global coverage of exchange of breeding animals. Typical country report excerpts were used to illustrate trends.

1.3 Scope The objectives were approached by conducting a separate global study and case study for each of the selected species.

1.3.1 Global study

The global studies attempted to understand the historic development and current status of global gene flow in each species. They considered the influence of domestication, breed formation, human migration, breeding methods, technological developments and global mobility on gene flow. Specific breeds were chosen to highlight typical developments of each species. References to other breeds were only made where appropriate. In cattle, breeds with Bos taurus ancestry and the Holstein and Simmental breeds were used to compare the typical dairy cattle situation with dual-purpose cattle. The Sahiwal and Brahman breed were chosen to depict the development of Bos indicus breeds due to their importance in the tropic ecozone. Ovis aries was represented by examples from Merino and Hair sheep breeds. The Merino breed was chosen as the outstanding historical example for the global spread of a breed due to the global demand for uniform fine wool fleeces. The double coated hair sheep breeds on the other hand serve as an example of breeds which are the result of thousands of years of adapting to their extreme production environments in which other sheep breeds rarely survive. Globally important examples of descendents of the Capra hircus goat breed are the Boer meat goat and the Angora fibre-producing breed. Sus scrofa subspecies are represented by the Large White and Duroc breed. The Large White includes very diverse types induced by its particularly high adaptability to wide- ranging conditions, including temperate as well as tropical regions. The Duroc is a meat- type pig with excellent adaptability and good combination characteristics, which has led to its wide spread to the tropics replacing local populations. INTRODUCTION 17

The current status of gene flow was indicated by exports and imports of genetic material, and influences of foreign genetic material on existing breeds. To represent global coverage, results for regional clusters were based on countries that were identified as typical for each region. Where information on representative countries was not available, this was noted along with the available information.

1.3.2 Case studies

Four selected case studies analysed in depth the typical aspects and effects of gene flow. Case studies were selected where such aspects could be shown and where regional information was available. The case studies are conducted from the perspective of tropical and subtropical countries. The improved Awassi and Assaf breeds from Israel were used to depict the historic development and current status of gene flow of a breed improved by within-breed selection. It follows up population developments in destination countries to conclude on their global impact. The goat case study depicts the historic global development of the Anglo Nubian breed and focuses on the impact of introducing a high performance breed to extensive smallholdings, specifically smallholder farms in Bolivia. The cattle case study depicts the origin, global transfer and utilisation of the improved local breeds Boran and Tuli. It focuses on its transfer to developed countries, the role of breeding organisations and the issue of access and benefit sharing. The pig case study depicts the impact of imported high performance pig breeds in Vietnam on socio-economic development and biodiversity.

18 GENERAL FEATURES OF GENE FLOW

2 GENERAL FEATURES OF GENE FLOW

2.1 Gene flow during domestication and breed formation Animals were first transferred by humans during domestication. Figure 1 gives the initial distribution of wild ancestors of domesticated mammal species.

Figure 1: Distribution of wild ancestors of major domestic mammals in Western Asia

(a) cattle, (b) sheep, (c) goats, (d) pigs, (e) overlap of ranges of cattle, sheep, goats, and pigs

Source: adapted from Reed (1984)

The far-reaching changes evoked by men between wild and domesticated species are the defined properties of domestication, i.e. taming, breeding in captivity, selecting for certain traits. Domestication is generally considered to have taken place in Neolithic times, with the primary centre in Western Asia. Domestication time frames and centres vary significantly from one source to another. Sheep are believed to have been domesticated about 9000 - 7500 BC, with the original centre in the Aralo-Caspian steppe (Zeuner, 1963). A second centre probably came up around 7000 - 6000 BC in Greece and East Europe (Ludwig, 1997). However, the origin of domesticated sheep is not fully clear. The goat was domesticated well before 7000 BC, although there is no solid agreement on their origin. The slopes of the Zagros mountains on the borders of present day Iran and Iraq seem to play a central role (Mason, 1981). From the west Asian domestication area, goat husbandry was spread out to west and east by the nomadic and semi-nomadic pastoralists and the sedentary (Nozawa, 1983). The first domesticated cattle appear on the Southern Anatolian plateau in Turkey about 6400 BC, and in Greece and Macedonia around the same time (Payne and Hodges, 1997), GENERAL FEATURES OF GENE FLOW 19 after sheep and goats. It is believed that cattle were domesticated independently in at least two distinct centres, one in western Asia (Bos taurus and Bos indicus) and the other in Southeast Asia (Bos (bibos) spp.). Conflicting theories exist about the centres of pig domestication. The earliest remains of domesticated pigs were found in Southeast Anatolia and were dated 7000 BC (Epstein and Bichard, 1984). Table 1 summarises references on first evidences of different domesticated species.

Table 1: Earliest evidence of domestication

Species Subspecies Time (BC) Location Country Sheep Ovis orientalis 8900 Zawi Chemi NW Iran3 Shanidar Goats Capra hircus 8000 Tepe Asiab NW Iran3 Cattle Bos primigenius 6400 Catal Huyuk S. Anatolia, Turkey3 Pigs Sus scrofa 65002 Qual’at Jarmo N. Iraq4 Sus scrofa About 5000 Hemudu , SE chirodonta China1

Sources: Zhimin (1991)1; Bökönyi (1974)2; Clutton-Brock (1981); Payne and Hodges (1997)3; Zeuner (1963)4; Zhong (1976)

As we assume that domestication was not a one-time-one-place event, gene flow would have been between different places of early domestication efforts and also between the domestic animals and the wild ones (Payne and Hodges, 1997; Payne and Wilson, 1999). The genetic make-up of each and every breed or population depends largely upon the genetic make-up of its founder group. This foundation group in turn depends upon the selection pressures it had encountered and the genetic make-up of its own founder group. As tribes of people migrated across the globe, they took samples of their livestock with them to their new homes. In each location the people and their livestock would adapt to their new environment through , either by the survival of those individuals genetically suited or an in-built ability to adapt to that environment. A sample of this population would then be taken with the next human migration to be the founders of a new community in a slightly different situation. The migration of people and livestock was not generally in one continuous direction. In most regions livestock was fairly continuously exchanged between communities. When breeds arose, through mutation or trade, with better survival or production characteristics than those in the local population, more of their progeny survived and the enhanced characteristics became common or even fixed within the group. Thus, other than geographically isolated situations, a gradual inflow of genes has modified every population that exists today (Henson, 1992). In the course of domestication and breed formation, a vast number of variations have developed to meet the many human needs over the millenniums so far. The third global inventory of domesticated breeds by the FAO found 6,379 breeds of 30 species (Scherf, 2000). However, in vast regions of the world many undefined breeds still exist, of which some might not be considered a distinct breed but still with broad characteristic variations. 20 GENERAL FEATURES OF GENE FLOW

Knowledge about breed development - at least for prehistorical times - is highly fragmentary. Nevertheless bone remnants and cultural-historic documents allow us to make some assumptions about early breed formation. The existence of is assumed in all early advanced civilisations like the Egyptian, Roman and Chinese civilisations, but also in Mediaeval Europe. For further details of early breeding history refer to Röhrs (1994). Prehistoric breeds were developed through artificial selection strongly accompanied by natural selection. Local breeds were particularly adapted to their specific natural environmental conditions. Moreover, so-called culture breeds may have developed before the industrial age, favourably predisposing breeds for the cultural economic and environmental conditions. Small, independent breeding units run by single farmers, landholders, or local communities contributed to genetic variation. Unintentionally this kept genetic resources safe.

2.2 Influence of breeding methods in further development and spread of breeds Two apparent breeding methods can be distinguished: (1) breeding with little investments and (2) breeding with increasing investments. Figure 2 depicts the historical development of breeding methods and their impact on gene flow.

Figure 2: Historical development of breeding methods and impact on gene flow

Time Prehistory Modern times

Breeding Spontaneous breeding with Planned breeding with methods little investments increasing investments

Influence on Rather little Increasingly strong gene flow “gene resource enhancing” “gene resource depleting”

Gene flow

Spontaneous breeding was predominant in prehistory - and probably partly overlapped with domestication. Spontaneous breeding assumes that the has some idea of a breeding goal, but does not have a definite plan to pursue it. The breeder is characterised by the so-called “breeder’s eye”, requiring only natural mating and subjective perceptions of phenotypes. The breeds produced by early breeding methods remained confined to the breeding unit and its immediate vicinity, unless caught up in early animal movements. This breeding “method” gradually disappeared when planned breeding efforts grew along with industrialisation. These new approaches are generally ascribed to British breeders in the middle of the 18th century. Breeding goals were defined and decisions based on the breeding goal and the mating procedure were applied. The British pedigree breeders approach typically involved little technical input, instead choosing monofactorial traits as the depending on show ring or sale performance, and then breeding to type. Tight upgrading rules such as closed herdbooks or many generations of upgrading into the herdbook narrowed the genetic variability. The early work of the British breed societies standardised breeds using strong formal criteria. Consequently, an increase of the frequency of desired alleles could be achieved, but restricted pedigree registration rules led GENERAL FEATURES OF GENE FLOW 21 to narrowing gene ressources with impact on the whole New and English speaking world. Through the sale of registered pedigree animals from the middle of the 19th century, gene flow became more and more commercial. In the long period of the 19th and 20th century breed associations evolved and a legislation and governmental involvement in breeding developed based entirely on phenotypic traits and mass selection. Since the beginning of the 20th century, plan-supported breeding developed with the accumulation of scientific insights. Selection programmes included objective animal performance recording, computing population genetic programmes, reproductive biotechnology and genome analysis. The result was increased investments. Commercial gene flow has made use of semen since the 1960s, embryos since the 1980s, and sexed embryos since the mid 1990s. Where profit allows highly developed breeding methods, gene resource depletion tends to be greater, for example in cattle and pigs, and less so for beef cattle, meat and wool sheep and meat and dairy goats. Complementing selectively developing breeds, systematic crossbreeding methods have developed to incorporate improved breeding material with, or even replace, original breeds. In developing countries, the biological potential is characterised by large numbers of animals with low production. Natural selection for fitness and hardiness reflects the sociologically founded principle of increasing flock size rather than productivity per animal. Therefore, most farm animals in the tropics and subtropics are diverse unimproved local landraces characterised by low production and high adaptability to adverse environmental conditions (Horst, 1999a). Additionally, need for cost-extensive and subsistence production systems in developing countries create totally different conditions for breeding strategies and breeding organisation. Developing a breeding strategy generally involves two phases. The first identifies selection criteria according to phenotypic, genetic and economic relevance, and estimates breeding values, and develops and implements breeding programmes. The second organises breeding structure, tests performance and manages data. For many reasons, applications of modern breeding methods in the tropics and subtropics are limited. Horst (1999b) and Gall (1999) give a detailed summary for these reasons. In developing selection programmes, for instance, identifying selection criteria requires economic evaluation of traits and evaluating genetic potential under adverse production conditions. The success of selection programmes within local populations may be limited because of the low genetic potential of the breed. Breeding work in developing countries is often restricted by few or no breeding organisations due to lack of finances or skills. Therefore, systematic breeding activities in developing countries are often limited to institutional or larger private farms. Small herds of institutional farms have too few animals for meaningful comparisons, and the production environment may differ greatly compared to the traditionally managed flocks (Gatenby, 1986). Activities on station are influenced by withdrawal of governments from the sector. Generally, institutional farms lack continuity in their policies and staff, which is unacceptable in a purebreeding programme especially for animals with long generation intervals. Since efficient breeding work and performance testing in smallholder farming systems is even more difficult for most species, developing countries generally face considerable difficulties to improve populations and distribute breeding progress. 22 GENERAL FEATURES OF GENE FLOW

Due to these difficulties, crossbreeding programmes have become more popular. To improve breeds, globally available improved genetic material must be available to developing countries. However, the suitability of the imported breed for improving the local breed has to be carefully evaluated under local conditions. Systematic imports of improved genetic material includes the exchange of tropical and subtropical breeds which themselves have already been improved under similar climatic conditions. Here, examples exist from countries with well organised breeding structures such as Israel, South Africa and Australia. China, the Gulf States, Brazil, Central America, and the Caribbean Islands may also qualify. Importing improved exotic breeds of European or North American origin promises greater improvements of single traits due to the higher difference in performance potentials between the exotic and local breed. Disadvantages are the acclimatisation risk for the exotic breed and genotype x environmental interactions, particularly reproduction and fitness traits. After all, importing high-yielding breeds requires a high level of management including feed and hygiene, to avoid performance depression, reproduction disorders or even physical degeneration. Since these requirements cannot be met in many developing countries, imported exotic breeds are preferably used in combination crossing or commercial crossings, to control the preferred exotic gene level and make use of (Horst and Reh, 1999). Under improving production conditions, replacing the local breed by systematically increasing the exotic gene proportion is another breeding strategy. In general, the diversity and dynamic of possible crossbreeding systems provides sustainable breeding strategies adapted to the local conditions. Due to the underdeveloped breeding structure in developing countries, integrating breeding progress into production is often attempted by disseminating improved sires to local breeders. However, the sustainable success greatly depends on an organisational programme to stabilise the appropriate gene proportion and the availability and ability of the local farmers to carry out breeding work.

2.3 Influence of technology and global mobility on the dissemination of genetic material Since the middle of the 19th century, gene flow has become more and more commercial through the sale of registered pedigree animals. However, international trade in live animals has many limitations. Veterinary barriers prohibit the free access to breeding material, and the climate and diseases at destination are often not suitable for certain breeds. Animal welfare rules limit the distance of surface transportation. Transportation costs subdivide the world market, particularly for high-yielding dairy cattle breed exports, into three zones on basis of origin: Europe, North America and Oceania. Biotechnology in permits, among other things, the creation of several tradable breeding products. In practice, these include semen and embryos and to a lesser extent sexed semen and embryos. Other possibilities are oocytes, body cells, sections of genetic information and clones (Polge, 1985). Compared to live animals these products 1) have a long shelf life by deep-freezing, making production and delivery independent from each other, 2) can be more easily reproduced thus securing a defined genetic supply, 3) are simple and cheap to transport thus reaching distant markets 4) are able to overcome veterinary barriers more easily, and 5) overcome some environmental and animal disease problems at destination in the case of exotic breeds intended for crossbreeding. GENERAL FEATURES OF GENE FLOW 23

Improved communications, global mobility and international business integration combined with advances in biotechnology have helped to develop a genuine world market for animal breeding products. The spread of artificial insemination (AI) and to a lesser extent of embryo transfer (ET) have substantially contributed to increased transfer of genetic resources to geographically distant locations. Even if these technologies are uneconomical as routine operations, as is sometimes the case in beef cattle, sheep and goats, they are nevertheless helpful and are used for international transfers. In the case of pig semen, however, the unresolved problem of freezing imposes a constraint. Worldwide transparency in animal breeding is being improved through the standardising work of the International Committee for Animal Recording (ICAR) and the better comparability of breeding values developed by its subsidiary, Interbull. So far, the main emphasis has been on dairy cattle breeds of the temperate zone. For developing countries, new biotechnology and global mobility offers easier access to genetic products of exotic breeds and offers easier transfer of genetic progress in regions with poor infrastructure or underdeveloped breeding organisations. It starts with introducing improved sires from nucleus flocks into the field via AI and and occasionally may be supported by reproduction technologies such as ET and hormonal synchronisation.

2.4 Gene flow indicated by Country Reports The Country Reports on Animal Genetic Resources are official government reports endorsed by each country’s national government. The reports describe the animal genetic resources of a country, analyse and report on the state of these resources and capacities to manage them, draw lessons from past experiences and identify problems and priorities (FAO, 2001). FAO uses the country reports as an input to the first The State of the World’s Animal Genetic Resources for Food and Agriculture. 165 countries have agreed to compile a Country Report before the end of 2004. The following overview is based on excerpts relevant to gene flow through FAO commissioned consultants who analysed the country reports. Altogether 90 excerpts were made available to the authors by FAO until December 2004. Since the description of gene flow is not a primary objective of the country reports the content relevant to this study differs significantly between the reports available. Annexes A 7, A 8, A 49 and A 53 show the sources of the 90 Country Report excerpts. The 55 underlined country reports had gene flow relevant information used in the compilations arranged in the chapters of the single species. Interpreting this analysis should be done with caution, since the information is not from a standardised source and often the sample is not representative. The narrative background is printed in full length in Annexes A 7, A 8, A 49 and A 53. About a third of the country reports contained information to single aspects of gene flow. The quality of this information differs strongly between country reports, from simple used in the country to detailed tables stating the date and number of individual animals or portions of semen imported. Generally the information regarding gene flow into the countries is better and more extensive than that about indigenous genetic material leaving. In the excerpts available only a few negative effects of importing exotic genetic material are mentioned. Of the 24 GENERAL FEATURES OF GENE FLOW points mentioned, the most important are replacement of indigenous breeds and failure of the exotic stock to adapt to local climatic and disease conditions.

MAIN FINDINGS OF GLOBAL STUDY 25

3 MAIN FINDINGS OF GLOBAL STUDY

In the following chapters the main findings of the global study are summarised by species. The extended versions of the global study for sheep, goats, cattle and pigs are found in the Annex chapters 9.2 to 9.5. The global study facilitated an overview over gene flow for each species. It illustrated the historical development of gene flow from the point of domestication, the influence of human migration and the development of breeding methods, modern breeding technologies and greater mobility on changes in quantity and on impact of gene flow. Finally it shows the current situation as far as it can be reconstructed on the bases of quantitative information. Due to the complexity of global flows, model breeds and breed groups serve as a red line through the global study.

3.1 Global gene flow of sheep The objective of the global gene flow study in sheep was to understand the magnitude and direction of movement of sheep genetic resources, to analyse the main factors affecting sheep gene flow and to conclude on its impacts on biodiversity and economic development in the destination countries. Limitations of the study and need for further actions are derived. The study consists of two parts: historic development, and current gene flow. We chose Merino and Hair Sheep breeds to highlight typical developments. References to other sheep breeds are only made where appropriate. We sourced data from the Eurostat statistic database, Country Reports on the State of Animal Genetic Resources, and from reports and publications from breeding organisations and regional experts. We show the influence of domestication on sheep diversity by summarising breed formation, human migration, and breeding methods. The depiction of the current status of gene flow focuses on exports and imports and indicators for foreign genetic material following regional clusters. Historic development of gene flow The world sheep population is about one third in developed and two thirds in developing countries (FAO, 2004a). With at least 1,747 known sheep breeds in the world, sheep breeds are impressively diverse (DAD-IS, 2004). This diversification has been driven by factors affecting breed formation, the occurrence of the characteristic wool fleece, the occurrence of the fat tail and the adaptation to a very broad range of environmental conditions (Haring, 1984). Among all domestic animals, sheep show the strongest relation to their production environment. Tribal migrations during the Neolithic transition are regarded as the biggest catalyst of gene flow after domestication. Sheep with favourable characteristics were traded and replaced breeds in other regions. This development is clustered into the regions of Europe, Africa, Asia, the Americas, and Oceania. 26 MAIN FINDINGS OF GLOBAL STUDY

The global spread of the Merino breed is an example of successfully replacing existing breeds. Hair sheep breeds of subtropical and particularly tropical climates represent those which have successfully resisted attempts to replace them. The history of the Merino breed shows that international gene flow has contributed significantly to diversification, and is an important tool for reacting quickly to changes in breeding goals based on changing production conditions. International gene flow starts where an international demand for a breed or its specific character appears. A suitable outstanding gene resource is required, which is often the result of successful breeding via genetic improvement. The importance of organised breeding is reflected by past global demand for temperate sheep breeds, in particular British breeds. For developing countries, examples of successful and unsuccessful international gene flow show that where the new breed is suitable to prevailing production conditions, international gene flow contributes to diversity. It was also shown that intensifying breeding work in developing countries influences the demand for, and success of, international gene flow. It has become evident that global mobility has increased and routes of international gene flow are complicated. Genetic transfer technologies such as AI and ET are less important where sheep live on extensive systems and so receive little attention. AI in sheep has had an important impact where using fresh semen is practical, such as today’s dairy sheep breeding programmes. There is no doubt that new techniques, which produce high conception rates and inseminate less invasively, could have a major impact on spreading genetic improvement in the sheep industries of many countries (Simm, 1998). Current status of gene flow Genetic resource exchanges are now intensive and complicated across the globe. The re- orientation from wool to meat and the search for niche production has started gene flow into wool producing countries. Gene flow is regarded as an appropriate tool to respond to quick changes in breeding goals. The new market economies in Central and Eastern Europe show a similar development. The intention to improve local breeds starts gene flow in most parts of the world, including developing countries. Main countries of origin appear to be Australia, New Zealand, North America, the UK and France. In rarer cases gene flow starts from developing countries. Where gene flow takes place, new breeds contribute to diversity by adding to existing sheep breeds and by creating new breed groups and synthetic breeds. Genetic diversity is lost when local breeds are completely replaced by imported breeds. It is not possible to assess the global impact on the limited information available. Even assessments of single countries contradict each other. Breeding organisations Countries of origin of gene flow generally have improved breeds as the result of advanced breeding organisations. In rare cases, genetic resources with desired genes or traits - in sheep mainly prolificacy, disease resistance, and adaptability - exist independently of breeding organisation. Then gene flow also comes from developing countries (for example the prolific D’man sheep from Morocco, the prolific Javanese thin-tailed and Javanese fat- tailed sheep of Indonesia, and hair sheep breeds of various origin). Access to improved genetic material does not depend on breeding organisations. Both developed and developing countries are in demand for new breeds. Developing breeding MAIN FINDINGS OF GLOBAL STUDY 27 organisations in developing countries will not necessary reduce the demand for foreign breeds. Although breeding organisation influences the impact of gene flow in the destination country, it was not possible to identify a relation between organisation and impact. Regulations Breeding regulations direct and control gene flow. High hygiene standards increase the opportunity to export genetic resources from a country, and to globally market them. At the same time, because less other countries will have similar hygiene standards, access to foreign genetic resources is reduced. Note that strict import and export regulations may lead to counter-productive illegal transfers of animals. Foreign aid and development projects Foreign aid and development projects can start or stop gene flow, depending on whether the aim is to replace indigenous stock or to genetically improve local breeds. If the suitability of imported stock for the production systems was taken into account, development programmes imported tropical and subtropical genetic material rather than temperate (for example developing and spreading the Dorper breed, and importing improved tropical hair sheep breeds into Southeast Asia). Research and commercial interest in specific genes With more possibilities for genetic change and relatively easy imports into new populations, identifying major genes is a main aim for both research and commerce. Breeds from countries with traditional sheep production systems move into the centre of interest as their genetic diversity is greater than in countries with intensive sheep farming. It has been shown that the interest in identified breeds goes far beyond their national importance (for example the Indian Garole breed, the Barbados Blackbelly from the Americas, the Moroccan D’man or Greek Chios sheep). Suitability of breeds for prevailing production systems If introduced breeds are suitable for the prevailing environments and production systems, gene flow contributes significantly to improving productivity of sheep farming. At the same time, sustainably establishing suitable foreign genotypes in local populations threatens local breeds either directly or through market competition. Compared to cattle and pigs, there are more sheep breeds, less extinct breeds and less fat-tail-risk breeds. Only goats have lower extinction rates (3%). However sheep breed variety is deteriorating, as the extinction rate since 1995 has more than doubled and the percentage of breeds not at risk has decreased. The high numbers of breeds with unknown status and, at the same time, the increasing number of recorded breeds suggest these figures may not be certain. Impact on economic development and reducing poverty If introduced improved breeds are suitable for the prevailing environments and production systems, gene flow contributes significantly to the economic development and poverty reduction by improving productivity. Gene flow is particularly important for economic development where the low production potential of the local breeds prevents quick improvements through within-breed selection. Local sheep genetic resources are however important for resource-poor farmers who depend on local breeds, adapted to their production systems, for their livelihoods. Such 28 MAIN FINDINGS OF GLOBAL STUDY farmers are often left out of improvement processes, particularly in remote mountainous regions. If the new breeds add to the diversity of national populations, gene flow offers the ability to quickly adapt to new production possibilities. Particularly, in rapidly developing parts of the world, gene flow can enhance production to meet the population’s growing demand for sheep meat, especially if accompanied by a change from extensive to intensive production systems. Market competition of intensive production units with extensive production systems also threatens the livelihoods of smallholders. The negative effects for smallholders are more severe the more the production systems of exotic and native genetic resources differ. Gene flow may also contribute to producing for niche markets, permitting and even encouraging the use of accessible foreign breeds that have characteristics suitable for local markets. This can give farmers, particularly sheep farmers, a chance to survive when otherwise facing severe reduction or complete loss of profitability. On the other hand, native genetic resources may contribute as much to production for niche markets, and so helps to conserve national genetic resources. Gene flow is particularly important when populations are reducing, or after a rapid decline such as after wars. When effective breeding work is difficult, and the base to restore and develop future populations is lost, gene flow is the only chance to rebuild the stock by importing genetic material. If the imported breeds are not suitable for the prevailing production systems, the impact on population composition is theoretically self-limiting. However, the economic losses for smallholders can be severe but temporary unless development projects or governments continue to promote or subsidise unsuitable genotypes. The losses are caused by diluting or replacing the indigenous genetic resources suitable for the smallholder production system. Impacts on biodiversity The influx of foreign genotypes into existing breeds has always been important to develop and improve sheep breeds. In the history of sheep, gene flow has contributed significantly to diversity. At the same time, improved genetic resources have always interfered with local resources. Current sheep gene flow is dominated by relatively few breeds from developed countries. However, due to the relatively rare use of biotechnology in sheep, the little attention paid to sheep production and the strong relationship between sheep breeds and their production environments, the impact on biodiversity is considered less severe than in other species. The impact of this gene transfer on biodiversity mainly depends on the suitability of the breeds to their new production environment, and can be measured as changes to the national population composition. If the new breeds are suitable and add to existing genotypes, then biodiversity increases. If new genotypes replace existing breeds, biodiversity is lost. Whether new breeds co-existence with or replace native breeds depends on various factors. One is the prevalence of the production system which the introduced genotype is meant for. In developing countries, where extensive small holdings prevail, dilution or replacement is likely if the new genotypes are suitable, are available, and are adopted by the local breeders. More often, the introduced genotypes require more intensive production MAIN FINDINGS OF GLOBAL STUDY 29 environments and therefore add to the more extensively managed local breeds. However, local market competition from intensive production units can put traditional production systems under threat and with them the local genetic resources. The more commercial the sector, the greater the threat of lost biodiversity. Producing for niche markets using local breeds to serve specific local consumer preferences may reduce this loss. Careful production system analysis is required to predict the impact of gene flow on biodiversity. While unsuitable breeds can have several negative effects, as given above, after a long time of adaptation these new genotypes may contribute to biodiversity by developing new breeds or traits. In Latin America for example, introducing unsuitable wool sheep breeds from Spain led to the now native and well adapted Chiapas sheep. As well as the main stream of gene flow from developed countries, both developed and developing countries which control some unique genotypes also export genetic resources. Global transfer of these genotypes generally contributes to biodiversity, particularly if they are threatened in their native countries. Horst and Reh (1999) give examples for some “old” European breeds, increasingly under threat in their native country, that are being crossbred in the tropics. By making use of their typical adaptability to marginal production environments, valuable genetic resources from developed countries can be conserved, as well as economic development in developing countries by improving local breeds. Producing for local niche markets using foreign breeds with suitable attributes are also examples of the benefits of biodiversity from transferred unique genotypes. Shrestha (2005) even points at developing composite populations from endangered and established sheep breeds, which may result in financial benefits and simultaneously conserve domestic animal diversity. Limitations of the study The most limiting factor of the global sheep study was the information available for current gene flow. Only for Europe is there an available statistical data base on breeding stock transfer and changes in national population composition, the “Eurostat” database. The records allowed only limited interpretation for gene flow. First of all, records did not include breeds, so that interpreting the influence of single breeds was not possible. It was also unclear whether these records reflected true volumes of animal transfer, such as whether animal transfers for development projects were included. Data on the exchange of other breeding products, such as semen and embryos were very limited. To gain information of the desired quality covering the different regions of the world, six breeding organisations and 21 experts from 12 countries were contacted. Only three contacts replied, and no statistics were made available. Official national statistics on live animal transfers were available via the FAO, but these transfers were not exclusively for breeding. In some cases, country reports included necessary information. However, if country reports included sheep transfer data, the usual focus was on imports rather than exports. Exports could only be traced indirectly as imports into other countries. Therefore, an important information source of gene flow was generally left out. This was particularly negative for developing countries, where records of animal movements rarely exist and country reports, if they exist, rarely contain the desired details. Export data from developed countries could have supplied that missing data. This means there is little information about gene flow from developing countries, particularly from South to South movements which would have been interesting. 30 MAIN FINDINGS OF GLOBAL STUDY

Increased global mobility has made gene flow more complicated and single transactions impossible to follow up, which also put severe limits on the global study. The example of the Booroola gene shows how the transfer of a few animals had major impacts on global gene flow. The Awassi case study demonstrated the effort it takes to collect the necessary information to assemble a nearly complete picture of gene flow for one particular breed. The study as a whole therefore cannot quantify gene flow on a global scope. It was shown that gene flow is a very complex and dynamic process with many factors varying the possible impacts from case to case. The impact evaluation of global gene flow on economic development and biodiversity in chapters 5.2 and 5.3 therefore had very general and qualitative conclusions, and refrained from an overall evaluation of global gene flow particularly its impact on biodiversity. The study objectively analyses possible aspects of gene flow and is meant to provide scientifically based information for decision makers in the discussion on global gene flow and its impacts. The study is certainly not suitable to finish this discussion. Need for further action The discussion on gene flow can only be fruitful if the impact of gene flow can be suitably evaluated. It has been stated above that global evaluation may not be possible, and that global interpretations are possible only if serious impact studies in each country add to the global picture. With the limited data bases available, further actions are needed to improve the records kept about current gene flow. Improving data quality and quantity in developed and developing countries, particularly import and export figures, is not realistic in the near future. Therefore the focus should be to derive reliable data on national population composition changes, which calls for frequent census studies at breed genotype level. This approach, which for most countries is challenging enough, would avoid the time consuming tracing of complex gene flow routes and instead starts right at evaluating impacts on biodiversity and economic development. Additionally, many weaknesses of the current information base could be reduced, such as the different comprehension of the status of national genetic resources by different groups. Each country, based on its national evaluation, should subsequently make use of the identified control mechanisms of gene flow to or from the country to improve the situation. One has to be aware that indigenous genetic resources are a source of adaptability for specific environmental challenges. Conserving these resources secures the national as much as global adaptability to different future production conditions and secures the livelihoods of animal holders who depend on these resources. In developing countries biodiversity and poverty reduction are two main claims future animal production has to deal with. Concepts which serve both claims at the same time by preserving or even improving smallholder livelihoods in marginal areas based on their indigenous genetic resources, should be rated highly. From these concepts derive the need to further identify and evaluate the genetic resources to determine their genetic distance to other breeds, their genetic potential and performance in different production systems and their risk status. What is more, biodiversity impacts must be taken into account when framework conditions are evaluated. Exotic breeds are currently being introduced for various reasons. If the genetic potential of the local breed is low, economic development calls for introducing suitable exotic breeds. “Suitable” requires responsibly selecting imported breeds, comparing performance of MAIN FINDINGS OF GLOBAL STUDY 31 different pure- and crossbred genotypes, and evaluating natural, economic, social and cultural conditions. Conserving national genetic resources through crossbreeding systems can be considered, as well as the use of foreign breeds which are under threat at their country of origin. Such approaches should preferably be carried out on farm with participation. However, if conserving indigenous genetic resources conflicts with economic development and poverty reduction, which can be the case, conservation cannot be carried out at the expense of animal producers. Conservation then has to be treated separately from but alongside economic development. Each country has to be aware that unique genotypes are in global demand. Based on the depicted complexity of gene flow, we conclude that once a genetic resource has been transferred, in whatever quantity, it is difficult to follow up the routes of gene flow. Gene flows should therefore be individually considered and recorded from the beginning rather than retrospectively.

3.2 Global gene flow of goats The objective of the global gene flow study was to understand the magnitude and direction of movement of goat genetic resources, the main factors affecting it and its impact on biodiversity and economic development in the countries of destination. Limitations of the study and further actions are derived. The study consists of two parts: historical development, and current gene flow. Boer goats and Angora goats were chosen to highlight typical developments. Material came from the Eurostat statistic database, Country Reports on the State of Animal Genetic Resources, and reports and publications from breeding organisations and regional experts. Information on the current gene flow in goats was scarce and less comparable between countries than that of other domestic livestock species. Historical development The majority of goat breeds are found in the tropics and subtropics. Historically, gene flow in goats starts with human migration, and breeds were formed based on regional clusters of human populations. Breeding organisations in the 19th and 20th centuries in England started an intensive gene flow of purebred goats for pure- and crossbreeding, leading to a diversity of goat breeds all over the world. In many countries, crossbreeding created new breeds. The outstanding example is the development of dairy breeds of Central Europe. Even in within-breed selection strategies, occasional introductions maintained an intermittent gene flow. Other breeds were formed mainly by importing pure stock or absorptive crossing. Breeds were often improved using genetic resources from outside the breeding area. This has happened in most improved breeds, fibre and dairy breeds in particular. In the 20th century, new (AI with deep-frozen semen and ET) increased global mobility of breeding products, disseminating and establishing new goat breeds around the world but to a far lower extent compared to cattle and even sheep. Current status of gene flow In developed countries, gene flow is currently driven by two main factors. Firstly, rare peculiar breeds some with special characteristics are moved from their home areas for niche markets or pets, for example the African Dwarf goat in some European countries. 32 MAIN FINDINGS OF GLOBAL STUDY

And secondly, promoted awareness of the potential of , for fibre, milk or meat, combined with unreliable profits in other livestock has led to imports of improved breeding stock to increase national stocks. One example is the introduction of the South African Boer goat to New Zealand. The available information for gene flow into developing countries reports the transfer of breeding material of a few improved breeds such as Boer, Anglo Nubian and Saanen goats originally from countries with advanced breeding organisations. Occasionally, the imports come from other developing countries which already possess improved breeding material for example the transfer of Toggenburg and Saanen goats from Kenya to Tanzania. The following main factors affecting gene flow were identified: Breeding and trade organisations affecting gene flow Not all countries have breeders’ organisations that actively work to improve goat genetics, due to socio-economic conditions. Nevertheless, in countries where breeders’ organisations are absent individual breeders may import breeding stock, but they are often restricted by regulatory difficulties. Private firms specialising in international livestock trade may help greatly, but goats lack the critical mass of profits to attract international firms to engage in their trade. However, breeders’ organisations from Britain, USA, France, Australia, South Africa and Germany have occasionally been engaged in such transactions. Restriction of goat gene flow by animal health regulations In order to prevent the introduction of major contagious diseases, governments tend to restrict livestock imports, particularly from other continents. These restrictions give countries which are free of certain diseases, in general developed countries, an advantage for exports but may exclude affected countries, in general developing countries, as sources. Various examples are given where gene flow is restricted by health regulations. As the animal health situation is constantly changing, so do the regulations on movement and import of livestock. The trend is towards severer restrictions. Foreign aid and development co-operation Over the past decades an increasing number of programmes have been started by governments, local and international non-governmental organisations (NGOs) and international donor agencies to promote or incorporate goat production in smallholder households to enhance regional development and reduce poverty. Examples from publications and project reports are given to illustrate the gene transfers. The gene flow is characterised mainly by the transfer of improved meat and/or milk breeds from developed countries. In rare cases, improved breeding material is transferred from one developing country to another. Experience also shows that development projects that are supported by governmental policies have a greater impact on the spread of specific breeds. Suitability of goat breeds for prevailing production systems The general adaptability of the goat enables it to be reared in both highly intensive and extensive production systems. In highly intensive production systems, there are the high- yielding, highly prolific and regularly breeding dairy types. At the other end of the spectrum, one can find all over the world many low performing breeds who excel by adapting to harsh environments, resisting diseases, and thriving under low-input systems. MAIN FINDINGS OF GLOBAL STUDY 33

Examples were given of successes and failures of goat transfers, depending on their fit to environmental conditions and systems of goat keeping. Impact on economic development and poverty reduction Developing countries display a demand for new breeds to increase productivity of goat farming or establish new systems, mainly dairy. With new biotechnology and increased mobility, goat genetic resources have become readily available globally. If introduced improved breeds are suitable to the prevailing environments and production systems, gene flow contributes significantly to the economic development of intensive and extensive production systems by improving productivity, i.e. increasing milk and/or meat yields. The importance of gene flow for economic development is particularly evident where the low production potential of the local breeds prevent quick improvements through within-breed selection. However, the greater the success of the introduced breed is, the greater the threat of losing biodiversity in the national population. Extensive goat farms in remote mountainous regions are however often not included in improvement processes, so these resource-poor farmers depend on local genetic resources and their adaptation for their livelihoods. Gene flow offers the chance to quickly adapt to new production opportunities. Particularly, in rapidly developing parts of the world, gene flow can enhance production to meet the growing demand for red meat. Then the development often goes hand in hand with a change from extensive to intensive production systems. Market competition between these intensive production units and extensive production systems can exclude smallholders from access to markets. Gene flow may also contribute to developing niche market products permitting and even encouraging the use of foreign breeds that have specific characters suitable to local market preferences. An example is the production with Angora goats in countries like Denmark. Gene flow is particularly important when goat populations are reduced, for example after wars. When there is no opportunity for effective breeding and the basis for restorating and developing future populations is lost, imported genetic material is the only chance for rebuilding stock. The choice of genetic material may be local breeds of neighbouring countries or exotic high-yielding stock depending on the production conditions of the region. If imported breeds are not suitable to the prevailing production systems, the impact on population composition is theoretically self-determined. However, the economic losses for smallholders can be severe, particularly if the introduction of these new genotypes is promoted or even subsidised through development projects or governments. Impact on biodiversity The influx of foreign genotypes into existing breeds has always been an important component in developing and improving goat breeds, and gene flow has contributed significantly to goat diversification. The growing global need for improved productivity of animal production has lead to an increased demand for improved genotypes. With the development of biotechnology and global mobility since the second half of the last century, foreign genetic resources are globally readily available. Current gene flow is dominated by source countries with advanced breeding structures, mainly developed countries, and a 34 MAIN FINDINGS OF GLOBAL STUDY small number of high performance breeds. However, due to the little attention paid to goat farming, the impact of global gene flow on biodiversity is less severe compared to other farm animal species. In general, the impact of gene flow on biodiversity mainly depends on the suitability of the breeds to their new production environments. If they are suitable and the new breeds add to existing genotypes in a controlled way, then biodiversity is increased. If new genotypes replace existing breeds, biodiversity is lost. Examples of both situations are well known. Whether replacement takes place after improved breeds are introduced depends on various factors. One is the prevalence of the production system which the introduced genotype is meant for. In developing countries, where extensive smallholder goat farms prevail, replacement mainly through crossbreeding are likely when the new genotypes are suitable and available. More often, these introduced genotypes require more intensive production environments and therefore may be established side by side with the more extensively managed local breeds. Careful production system analysis will be needed to predict the impact of gene flow on biodiversity. If the introduced breeds are not suitable, the impact on biodiversity is theoretically limited. However, if the introduction is promoted or even subsidised through development projects or governments, or if indiscriminate crossing takes place particularly over time and in considerable quantities, existing breeds can deteriorate. However, over a long time of adaptation, these new genotypes may also contribute to biodiversity through developing of new breeds or traits. Apart from the main stream of gene flow of high-yielding breeds from developed countries, gene flow takes place from developed and developing countries, which control some unique genotypes (Boer, Nubians). Many goat breeds in developed countries have been formed from foreign genetic material (e.g. Anglo Nubians). Limitations of the study The most limiting factor of the study was the information available for current gene flow in goats. Only in Europe was a statistic database (Eurostat) available and even there, the records allowed only limited interpretation with regard to gene flow. First of all, records were not made at breed level, so the influence of single breeds could not be followed. It was also unclear how the records reflected true volumes of animal transfer, and whether development project transfers were included. Data on the exchange of other breeding products, such as semen and embryos, were in general lacking. Country Report excerpts are less extensive regarding goat transfers compared to other species, and generally focus on exports rather than imports. The content varied greatly between the available reports and many excerpts did not mention goats at all. This made it particularly hard to find information on gene flow into developing countries. Here records of animal movements are rare and country reports do not contain the desired details. The latter limitation shows the lack of information of gene flow from developing countries, particularly from South to South, which would have been interesting to examine. With this extremely limited information on current goat gene flow, the study had to refrain from quantifying gene flow globally. It is known that gene flow is a very complex and dynamic process with many factors varying the possible impacts from case to case. The impact evaluation of global gene flow on economic development and biodiversity, as MAIN FINDINGS OF GLOBAL STUDY 35 treated in chapter 5.2 and 5.3, therefore needed to conclude very generally and theoretically. Need for further action The discussion on gene flow can only be fruitful if its impact can be suitably evaluated. It has been stated above that global evaluation is doubtful and that global interpretations become possible only if serious impact studies are available. Further actions are needed to improve the information status on current gene flow. It is unlikely that data quality and quantity in developed or developing countries, particularly import and export figures will improve significantly in the near future. So the focus should be to derive reliable data on national population composition changes, which calls for regular census studies of breeds and genotypes. This approach, which for most countries is a challenging enough task, would avoid the time consuming tracing of complex gene flow routes and instead starts right at evaluating impact. Additionally, many weaknesses could be minimised, such as the different values put on national genetic resources by different groups. Each country should subsequently make use of the identified control mechanisms of gene flow to or from the country. One has to be aware that indigenous genetic resources are a source of adaptability for specific environmental challenges. Conserving these resources secures the livelihoods of animal holders who depend on these resources. In developing countries biodiversity and poverty reduction are two main - often opposite - claims which future animal production has to deal with. Concepts which serve both claims at the same time by preserving or even improving smallholder livelihoods in marginal areas based on their indigenous genetic resources, should be valued very highly. From these concepts derive the need to identify and evaluate goat genetic resources for genetic distance to other breeds, their genetic potential and performance in different production systems and their risk status. What is more, there is also the need to take biodiversity impacts into account when framework conditions are evaluated. Improved breeds are currently being transferred for various reasons. To claim suitability, imported breeds must be selected by evaluating performance of different pure- and crossbred genotypes against the natural, economic, social and cultural framework conditions. Such approaches should preferably be carried out on a farm with farmer participation. However, if conserving indigenous genetic resources and economic development and poverty reduction contradict each other, conservation efforts cannot be carried out at the expenses of animal producers. Conservation then must be treated separately but parallel to economic development. Each country has to be aware that unique genotypes may be in global demand. Based on the depicted complexity of gene flow, we conclude that it is difficult to find the routes of gene flow after the event, in whatever quantity. Gene flow should therefore be individually considered and recording practices put in place in advance, rather than retrospectively attempting to measure the transfer.

3.3 Global gene flow of cattle The objectives of the global gene flow study in cattle were to understand the magnitude and direction of movement of cattle genetic resources, to analyse the main factors that affect cattle gene flow and to determine its impacts on biodiversity and economic 36 MAIN FINDINGS OF GLOBAL STUDY development in the destination countries. Limitations of the study and necessary further actions are derived. The study consists of two parts: the historic development and the current status of gene flow. As typical examples for Bos taurus, we chose the Holstein and Simmental breeds, and for Bos indicus we chose the Brahman and Sahiwal breeds. Data was sourced from the Eurostat statistic database, Country Reports on the State of Animal Genetic Resources, and from reports and publications from regional experts and breeding organisations, especially the breeding organisations of Simmental, Holstein and breeds. We summarise how historically domestication, breed formation, human migration, and breeding methods have influenced cattle diversity. The chapter on current gene flow summarises export and import data and indicators for foreign genetic material introduced to clusters of regions. Historic development Historic gene flow shows that domestic cattle ancestors and their dissemination in the old and new world contributed to the development and diversity of current domestic cattle breeds. During the Neolithic transition, migrating tribes took along their animals and so affected gene flow in cattle early in history. Occasionally, migration also made wild cattle extinct. In the course of time, local strains developed through natural selection that were adapted to their environments. In the new world, settlers brought cattle with them to create living conditions similar to their homeland and to insure against food deficiencies. A source of significant gene flow is cattle trading. This is demonstrated by centuries of established trading houses, empires, and long trade routes by land and sea. Breed societies were formed in 18th century Britain, and pedigree breeding spread to continental Europe and the New and English speaking worlds. This decreased genetic variety by forming breeds on formal criteria. Early breed selection focused on conformation, and productivity advances were limited. Only later did selection for performance and genetic evaluation lead to more rapid productivity increases. Cattle selection is based on utility or market requirements and the climate. The influence of the Black and White breed on dairy cattle in the temperate zone is considered the most significant event in cattle gene flow. It absorbed and displaced other temperate zone dairy and dual-purpose breeds, heavily impacted gene flow and reduced genetic resources. At the same time, the few top and top lines decreased genetic variety within the breed. Successfully expanding purebred Holstein into the dry tropics and subtropics and failures by introducing them to the humid tropics showed that gene flow can establish breeds or breed groups when suited to prevailing production requirements. As many efforts to establish purebred temperate dairy breeds in the humid tropics have failed, dairy breeders there now focus on forming composite taurine-zebuine cattle for milk. The global influence of the Brahman and Sahiwal breeds is reviewed in this context. The spread of the Simmental breed is an example of a globally influential dual-purpose breed, mainly for beef. Gene flow of specialised beef breeds has been less intense than dairy or dual-purpose breeds, as reproduction is mainly by natural service and there is not yet sufficient knowledge about the comparative efficiency of breeds kept on natural pasture. However, as MAIN FINDINGS OF GLOBAL STUDY 37 an exception of the rule considerable numbres of Brahman and Hereford have been spread around the world. Major restrictions to the international live animal gene flow are identified. However, successful use of biotechnology and global mobility has proved to be more striking in cattle than other species. The international gene flow in cattle increased substantially after the technique of deep-freezing bull semen was widely adopted in the 1960s. Techniques to deep-freeze bovine embryos from 1980s onwards never attained comparable importance as a vehicle for gene flow. Major differences in the commonplace use of this biotechnology are seen between developed and developing countries and between dairy cattle, dual- purpose, and beef cattle breeds. Current status First, information on selected cattle exports and imports of live animals and semen were compiled for the regional clusters Europe, North America, South America, Asia and the Middle East. For Europe, information was sourced from the Eurostat statistic database. For other regions, no database was accessible and information was taken from publications of breeding organisations and from communication with regional experts. Gene flow is currently a considerable global exchange of live animals and semen. The main origins are developed countries, with North America and Europe as major players due to their advanced breeding and trading organisations. International gene flow starts due to quick reactions to market opportunities and changing production purposes, and the general need to improve local genetic resources. Destinations of gene flow are therefore developed and developing countries alike. The largest inflows are reported across Europe, followed by Western Asia. Exchanges of breeding stock or semen vary in quantity and impact. The current dairy cattle semen trade is the most intensive and is quite different compared to other livestock species. Examples are given of impacts of international gene flow on national breed composition, and on introducing foreign genetic material into existing breeds. However, the impact study could not quantify the effects, as the information base was extremely limited. Common impacts were: reduced proportions of existing breeds in national cattle composition in favour of imported breeds, and increased foreign blood shares in existing breeds. In some cases, continuous gene flow into local breeds finally replaced them. According to information from selected Country Reports, many breeds were being replaced in Western and Northern Europe, some replaced in Africa, Eastern and Southern Europe and Latin America. The new breeds were Holstein, Simmental and Brahman. Breeding and trade organisations The current situation in cattle semen trade - particularly dairy cattle semen - is fostered through the high level of breeding and trade organisation in developed countries, which gives them an indiscriminate advantage over developing countries with regard to exploitation, and dissemination of their genetic resources. At the same time, cattle breeding and trade organisations have experienced a remarkably strong concentration process in the last decades. The international exchange of genetic material from a decreasing number of sires has resulted in an increasing loss of genetic variation with global impacts for developed and developing countries. The in general lower AI coverage in developing countries causes disadvantages also in the access to exogenous gene flow although with 38 MAIN FINDINGS OF GLOBAL STUDY considerable variability between countries and regions. As AI coverage in developing countries is increasing, gene flow of temperate breeds increases parallely. Regulations Cattle breeding products are mainly transferred by a few international breeding companies, making transport of any magnitude and distance technically possible. As live animals transfers face restrictions by veterinary regulations and welfare considerations, the great success of biotechnology in cattle has created tremendous possibilities in international gene flow. In general, high hygiene standards of a country permit almost unlimited exports to all parts of the world, while it prevents stock being introduced from countries with lower hygiene standards. The emergence of BSE and the outbreak of Foot-and-Mouth disease in Europe in the late 1990s for example have influenced the exchange of breeding animals, but also semen and embryos for years. In the cattle sector, government policies have been the main influence on breeding and gene flow. Regulations in trade and breeding legislation of cattle are generally market- regulating subsidies. Other government activities influence gene flow by importing selected breeds as part of national plans or projects. Research and commercial interests in specific genes Interest in specific genes affects gene flow by prompting the transfer of these genes from their countries of origin to those places where the commercial or research interest exists. In cattle, interests lie in identifying gene markers to reduce production costs, improve product quality, marbling of beef, parasite resistance and feed efficiency. Foreign aid and development projects Cattle, particularly dairy cattle, have always attracted more interest than other species. Subsequently, foreign aid and development projects are numerous in this sector. A widespread attempt to increase the national output from animal production in developing countries was to introduce improved temperate breeds for pure- and crossbreeding, so starting gene flow considerably larger than in other species. However, due to complex reasons such as genotype-environment interactions and socio-economic prevalence, success varied. The introduced breeds must be suited to the climate for projects to be successful, and for the impact of the gene flow to be sustained. Looking at projects that introduced improved breeds into the most difficult production environment, the humid tropics, we can see the correlation between suitability and success. Additionally, stable politics and socio- economic factors have been identified as important for a project’s success. Suitability of breeds for prevailing production systems The suitability of imported breeds for prevailing production systems determines their impact on existing breeds. In cattle more than in other species, the prevailing production conditions, especially the climate, decide over the breeding or production purpose. The spread of the Holstein breed is a vivid example of the global impact of a single- purpose high-yielding dairy breed suitable for the whole range of temperate, dry tropical and subtropical zones where the production conditions favour dairy production. Efforts to introduce Holstein cattle into the humid tropics have generally failed. MAIN FINDINGS OF GLOBAL STUDY 39

Many breeds are suitable for diverse purposes in the different ecozones: While B. taurus breeds from Britain and from the European continent are suitable for the temperate zone, B. taurus x B. indicus crosses, Zebu breeds, Sanga breeds and other indigenous breeds are suitable for the hotter areas. Productive dairy cattle for the humid tropics require some taurine blood with a high gene frequency for milk combined with a Zebu breed for adaptability. In the tropical and subtropical zone of Asia breeding emphasises crossbreeding for milk production but there is also some development of beef. When cattle producers need multi-purpose cattle or if product prices do not justify costly inputs, other breeds and breeding methods are used. Here, the Sahiwal (Zebu) and its crosses with temperate zone dairy and dual-purpose breeds are useful. The Simmental breed was originally dual-purpose and has proved its suitability under various production systems and climates. The use varies from country to country depending on milk prices and the milk-beef-price ratio. In Central and Eastern Europe the breed is used as a genuine dual-purpose breed. In the New World and Northwest Europe it is mainly used for beef. Simmental genetic material has been used in crossbreeding for industrial crosses, to develop new composite breeds, and to upgrade local cattle breeds. Simbra became the most widely distributed synthetic breed to which Simmental has made substantial contributions. The Brahman, adapted to hot climates due to its superior thermoregulation, and the Santa Gertrudis breeds had the most influence on beef cattle breeding in the tropical and subtropical zones of the New World. But they were found to be poor subsistence milk suppliers for pastoral herdsmen. The lack of genotype-environment interaction evaluations led to the foundation of the International Bull Evaluation Service (Interbull) to avoid sires of temperate origin being genetically overestimated when used in tropical and subtropical countries. Impact on economic development and poverty reduction Developing countries are in demand for improved dairy and multiple-purpose breeds to increase productivity of farming. Based on the good global access to cattle genetic resources, gene flow of dairy cattle breeds is the most extensive global exchange of genetic resources compared to other cattle breeds and pigs, sheep and goats. If the introduced improved breeds are suited to prevailing environments and production conditions, gene flow contributes significantly to developing the economy and reducing poverty by improving productivity and thus milk and red meat production. Gene flow is particularly important for economic development where the low production potential of the local breeds does not allow quick improvements using within-breed selection. Local genetic resources are however important for resource-poor farmers who depend on them and their adaptation to their production systems for their livelihoods, as they are often not included in improvements. Impact on biodiversity Introducing foreign genotypes into existing breeds has always been important to develop and improve cattle breeds. In the history of cattle, gene flow has contributed significantly to diversity. The growing global need for improved animal productivity has increased demand for improved genotypes. Based on the current good global access to cattle genetic resources, current cattle gene flow is dominated by source countries with advanced 40 MAIN FINDINGS OF GLOBAL STUDY breeding structures, mainly developed countries, with the main gene flows north to north and north to south. At the same time, animals from developed countries increasingly belong to a small number of high performance breeds selected over the last two centuries and strongly influenced by controlled science-funded breeding programs. They give high yields but require standard conditions and high inputs to exploit their potential. The global gene flow of relatively few numbers of breeds is significant and most evident in dairy cattle breeding. The impact of this gene transfer on biodiversity depends mainly on the suitability of the breeds to their new production environment, and is measurable in changes to the national population composition. If the breeds are suitable and the new breeds add to existing genotypes, then biodiversity increases. If the new genotypes replace existing breeds, biodiversity is lost. Examples for both situations are numerous in cattle transfer. Whether the new breeds co-exist with or replace native breeds depends on various factors, such as the prevalence of the production system, which the new breed is meant for. In developing countries, where in general extensive smallholder production systems prevail, dilution or replacement, mainly through crossbreeding, are likely where the new genotypes are suitable for, available to and adopted by the local breeders. Then, economic development of the smallholders’ livelihoods may conflict with increasing biodiversity. More often, these introduced genotypes require intensive production environments and therefore add to the extensively managed local breeds. However, through local market competition intensive production units threaten traditional production systems - the livelihoods of smallholder farmers - and with them the local genetic resources. The more commercial the sector is the greater the threat of lost biodiversity. Therefore, dairy cattle breed diversity possibly faces the most severe threat compared to other cattle breeds or small ruminants. Niche market production using local breeds to serve specific local consumer preferences that may preserve diversity is less important in cattle production compared to sheep and goats. Apart from the main stream of high-yielding breeds from developed countries, gene flow takes place from both developed and developing countries, which control some unique genotypes. Global transfer of these genotypes generally contributes to biodiversity, particularly if they are threatened in their native countries. Limitations of the study The information base of current cattle gene flow, particularly dairy cattle, was much better compared to the other species. Nevertheless, it was limited. Countries with highly organised cattle breeding structures were expected to have statistical databases on breeding stock transfer and changes in national population composition. However, only Europe had such records available and accessible in the form of the Eurostat statistic database. Furthermore, the records allowed only limited interpretation of gene flow. Records did not include breed details so the influence of single breeds could not be determined. It was also unclear how these records reflected true volumes of animal transfer, for example whether animal transfers for development projects were included. Only the exchange of purebreeding stock was recorded. With commercialisation, export and import statistics are inadequate to depict gene flow, and company figures would be more suitable. As international breeding companies restrict access to their information, this information was in general limited. MAIN FINDINGS OF GLOBAL STUDY 41

The FAO gave official national statistics on the exchange of live animals but the exchanges were not exclusively for breeding. In some cases country reports included the necessary information, but when they did the usual focus was on imports not exports. Exports could only be traced indirectly as imports into other countries. Therefore, an important source of recorded information on gene flow was generally left out. This was particularly negative for developing countries, where records of animal movements rarely exist and country reports, if they do exist, do not contain the desired details. Export data from developed countries could have supplied that missing data. This means there is little information about gene flow from developing countries particularly South to South movements, which would have been interesting. In general single transactions are impossible to follow up. However, there were examples for all species where the transfer of a few animals had major impacts on global gene flow. These examples become clear only retrospectively through changes to national population composition. This indicator of gene flow is an important tool in the study. One information source were the Country Reports on animal genetic resources. The information content varied greatly between the available reports. With increasing global mobility and biotechnology, cattle gene flow in particular has increased and become more complicated, putting severe limits on the global study. The study had to refrain from quantifying cattle gene flow on a global scale and from concluding on any other than a very general sense. Need for further action Considering the limited data available to this study, further actions are needed to improve the information available about current cattle gene flow. Due to the importance of international breeding and trade organisations in cattle gene flow, the quantity of information recorded was satisfactory. However, the accessibility and the quality of data must be improved considerably for satisfactory impact evaluation studies in future. Improving data quality and quantity, including import and export figures, in developing countries particularly, is not realistic in the near future. Therefore the focus should be to derive reliable data on national population composition changes, which calls for frequent census studies of breeds and genotypes. This approach, which for most countries is challenging enough, would avoid the time consuming tracing of complex gene flow routes and instead starts right at evaluating impacts on biodiversity and economic development. Additionally, many weaknesses of the current information base could be reduced such as the different comprehension of the status of national genetic resources by different groups. Each country should, after its evaluation of the national situation, control gene flow to improve the situation. In developing countries biodiversity and poverty reduction are two main aims animal production has to deal with. Concepts, which serve both at the same time, by preserving or even improving smallholder livelihoods in marginal areas based on their indigenous genetic resources, should be rated highly. Further identification and evaluation of genetic resources is required, to measure genetic distance to other breeds, their genetic potential and performance in different production systems and their risk status, although much work has already been done in cattle compared to other species. Biodiversity impacts should be taken into account when evaluating framework conditions. 42 MAIN FINDINGS OF GLOBAL STUDY

Developing suitable breeds requires responsibly selecting imported breeds, comparing performance of different pure- and crossbred genotypes, and evaluating the natural, economic, social and cultural conditions. Conservation has to be treated separately but alongside economic development. Conserving national genetic resources through crossbreeding can be considered as well as using foreign breeds, which are under threat at their country of origin. Such approaches should preferably be carried out on farm with farmer participation. Each country has to be aware that unique genotypes are in global demand. Given the complexity of gene flow, once a genetic resource has been transferred in whatever quantity it is difficult to follow up the routes for global exchanges. The control of gene flow should therefore be individually considered and recorded from the beginning and not retrospectively.

3.4 Global gene flow of pigs The objective of the global gene flow study in pigs was to understand the magnitude and direction of movement of pig genetic resources, to analyse main factors effecting pig gene flow and to conclude on its impacts on biodiversity and economic development in the countries of destination. Limitations of the study and need for further actions are derived. The study consists of two parts, the historic development and the current status of gene flow. The Large White and Duroc breed have been chosen to highlight typical developments. Material from various sources has been used. Data was sourced from the Eurostat statistic database for the exchange of purebred pigs, from Country Reports on the State of Animal Genetic Resources, and from reports and publications from regional experts and breeding organisations. Considered that crossbred animals have gained increasing importance, the trade with breeding pigs cannot be quantified easily. Commercial companies dominate the exchange market of some major countries and details are normally confidential. Information on the current status of gene flow was therefore largely based on interviews with Dr. Maurice Bichard, a senior expert and a retired leading scientist of the former Pig Improvement Company Limited (PIC), and Mr. David Steane who provided additional information for Southeast Asia. Isolated statistics of Southeast Asia and Canada were additionally used to fill information gaps. With regard to the historic development, influences of domestication, breed formation, human migration, and breeding methods on the diversification of pigs are summarised. The depiction of the current status of gene flow focuses on selected export and import information and on indicators for the introduction of foreign genetic material characterised for regional clusters. Historic development Although conflicting theories exist about the domestication of pigs, it is evident that domestication initiated historic gene flow in pigs and human migration and colonisation activities continued the global dissemination of pig breeds. The formation of breed societies in the 18th century in Britain pushed the development of pig breeds and distinct types based on pigs imported from China and Indo-China. MAIN FINDINGS OF GLOBAL STUDY 43

In the 1930s Mendelian genetics was combined with statistical methods to produce a theoretical basis for animal improvement. This methodology took gradually over from traditional methods as the driving force for pig improvement. Additionally, a general trend to more concentrated pig production was observed worldwide. Since the 1970s, profound changes in pig production influenced gene flow significantly. Technical changes led to an increased professionalism and through increased investment demands to an increased commercialisation. As the whole production chain increased in scale, organisational changes have been driven by increased competition. Gene flow since then has been characterised by a new dimension in internationalisation and liberalisation. The expression and expansion of breeding in pigs reflects on the degree of commercialisation in pig gene flow with increased tendencies to monopolisation. Breeding pigs worldwide are today mainly distributed by a few international breeding companies, particularly in the case of hybrid pigs. While in the past developed countries contributed most to the dissemination of genetic material, nowadays there are strong breeding enterprises also e.g. in Thailand, the Philippines and China. In the second half of the 20th century increased global mobility has intensified and complicated the routes of transfer of breeding stock. Freight costs, veterinary restrictions, welfare considerations and exchange and inflation rates are the main restriction. Although international breeding companies strongly utilise all biotechnological methods, until today, mainly live breeding pigs are transferred. In this development, source countries of gene flow were those which controlled some unique genotypes, and then those which emerged quickly from the old, conservative breed-society and government-dominated structures. The formation of commercial pig breeds in Europe and North America depended heavily on the introduction of foreign genetic material from Asia and soon after their establishment they replaced indigenous breeds. The other way round, Asian countries have imported pigs of European origin for crossbreeding. In the first four decades after World War II all countries in the Americas, in Japan, Taiwan, Korea, Thailand, and Singapore were strongly influenced by breeds from the USA. Similarly, purebred stock from Britain went to the old commonwealth countries Australia, New Zealand, South Africa, Kenya, and Zimbabwe. Current status Current gene flow in pigs is more than in other species driven by growing international competition of breeding companies. The complex organisation of international breeding companies complicates the routes of gene flow and makes it difficult to follow up impacts. The main planned gene flow takes place from Europe and North America into developing countries of the South. Export quantities are difficult to estimate, mainly because almost all companies are primarily focused on providing breeding stock for their home country. Generally commercial companies sell relatively few high-priced pigs to an overseas customer. As in recent years the market for breeding stock has become more difficult within their home countries, the national companies have taken increasing interest in exports. Apart from the main planned flows from Europe and North America into developing countries of the South, there is unplanned "leakage" or gene flow from primary customers to neighbouring countries. Although the main pig gene flow is expected to take place in commercial production, there is a continuous trend to replace local breeds with exotics. 44 MAIN FINDINGS OF GLOBAL STUDY

In the analysis of the historic development and the current status of gene flow three main factors affecting gene flow were identified: 1. Breeding organisations In pig breeding, the impact of changes in breeding and trade organisation on gene flow is most evident. With the involvement of “new” production areas in Latin America, Southeast Asia, and Eastern Europe and the exponential growth of the North American agro-industry in combination with trade liberalisation, a situation of increased competition and commercialisation occurred in the sector. The competition created the necessity to increased quality, diversity and quantity of genetic products. This development called for corresponding breeding programmes and the advanced use of biotechnology. Under the economic pressure of reduced profit margins, and to secure breeding progress and sale volumes, the concentration level rose. An increased tendency to monopolisation was observed with leading positions of those countries which most quickly had emerged from the old, conservative breed-society and government-dominated structures, giving Europe and North America unattainable advantages over the rest of the world. However, nowadays there are strong breeding enterprises e.g. in Thailand, the Philippines and China. In consequence, national pig breeding programmes as well as pig breeding companies are becoming increasingly similar in their structures. With regard to gene flow, the numbers of breeds used in the commercial pig breeding programmes is reduced to a few world wide and their genetic variation is considerably reduced. Increasing concentration has lead to the increased inflow of these breeds into many parts of the world. The development of hybrid breeding leads to the control of gene flow through the commercial companies. As in general main pig gene flow takes place in commercial production, there is a continuous trend of replacement of traditional breeds through the competition of commercial and traditional pig production in the national market. On the other hand, unique germplasm from Asian breeds plays a vital role in commercial pig breeding companies of European and North American origin. In conclusion, important source countries need the control over unique genotypes to ensure their market position. In pig breeding it becomes evident that with the increasing degree of commercialisation gene flow becomes less traceable through export and import. Single company transactions are impossible to follow up but may have high impacts. 2. Regulations As dissemination of breeding pigs is mainly facilitated by few international breeding companies particularly for hybrid pigs, technically transport of any magnitude and distance has become possible. Practically, gene flow faces restrictive challenges through veterinary regulations and welfare considerations as still mainly only breeding animals are transferred. Strict health regulations also force franchises of the big international pig breeding companies to utilise existing breeds. Strict health regulations have on the other hand prevented countries to benefit from international progress. 3. Research and commercial interests As in the pig sector, consumers preferences play an important role there is a potential demand for specific genes to serve these preferences. Some examples of Taihu imports by France, The Netherlands, the UK and USA fall under this category but no commercial MAIN FINDINGS OF GLOBAL STUDY 45 impact has been stated so far. Some interest in Vietnamese “pot-bellied” pigs as pets has been reported. Foreign aid and development projects In general, not much information was accessible on foreign aid and development projects and their impacts on the pig sector. In the case of Vietnam, as reported in detail in the case study, exotic breeds have been promoted and imported in the frame of governmental strategies on an intensive scale and the impact of these state-run breeding stations has been considerable with regard to gene flow. In national development projects and breed distribution projects the distribution has been executed. Additionally, the Vietnamese research and educational system supported the use of exotic breeds and crossbreds inclusively advanced management techniques and the promotion of the use of AI through the building of AI stations. Since the late 1980s with the start of the open door policy, additionally governmental and non-governmental organisations have played a role in introduction and distribution of higher-yielding pig genotypes in smaller and bigger projects. In consequence, most indigenous breeds show declining population trends and the majority of breeds is in a vulnerable or critical state, even facing extinction. Only a small number of breeding centres and research institutes keeps local breeds for crossbreeding and conservation. Due to the non-centralised breeding organisation in Vietnam, neither genetic improvement nor breed replacement and conservation are uniform across regions. Suitability of breeds for prevailing production systems In commercial pig production, the environment is normally shaped according to the needs of the pigs. Adaptability problems are therefore in most cases reduced. Therefore, in breed comparisons of indigenous with exotic breeds and their crossbreds, the exotic genotypes hold the leading rankings. Subsequently, commercial pig stock even in tropical and subtropical climates, principally exists of commercial hybrids but also exotic purebreeding lines. However, genotypes from temperate regions and their crosses face severe problems when confronted with traditional management conditions in developing countries. Furthermore to avoid loss of heterosis in consecutive generations continual inflow of exotic germplasm is necessary which makes the success of crossbreeding schemes at small-scale level doubtful. Institutional, organisational and technical support are required for these crossbreeding schemes not to fail. In case of no further access to exotic breeds, the impact of punctual crossbreeding on the local populations vanished from generation to generation. Impact on economic development and poverty reduction With regard to the high degree of commercialisation and competition of international pig production, the impact of gene flow on economic development is different from other farm animal sectors. Economic development of pig breeding and production is taking place where developing countries have managed to participate in the international competition, as for examples the case with strong breeding enterprises in Thailand, the Philippines and China. In commercial pig production, suitability of the imported genotypes to traditional management conditions in developing countries are of minor importance. As mentioned above the success of crossbreeding schemes at small-scale level are doubtful. Local genetic resources are therefore important for resource-poor farmers who depend on local breeds 46 MAIN FINDINGS OF GLOBAL STUDY and their adaptation to their production systems for their livelihoods. Through market competition of intensive production units with extensive production systems, the livelihoods of smallholders are under threat. As production systems of exotic and native genetic resources differ most and the degree of commercialisation is most advanced, negative effects for smallholders are the most severe in the pig sector. Apart from the main stream of gene flow of high-yielding breeds from developed countries, gene flow takes place from countries, which control some unique genotypes displaying specific characters with regard to local market preferences. Developing countries which are of control of a genetic resource of international demand have to be aware of the international competition and insure the national exploitation of the resource with regard to its economic potential. Impact on biodiversity The influx of foreign genotypes into existing breeds has always been an important component in the development and improvement of breeds. In the history of pigs, gene flow from various directions has contributed significantly to diversification. Current gene flow is domineered by the high degree of commercialisation in the pig sector. Transferred breeding pigs increasingly belong to a small number of high performance breeds and hybrids selected for high yields requiring standardised conditions and high inputs for exploitation of their potential developed over the last two centuries and strongly influenced by controlled scientifically funded breeding programs. These introduced genotypes require more intensive production environments and therefore add to the more extensively managed local breeds. However, through competitive effects on the local markets, intensive production units put traditional production systems under threat and with them the local genetic resources which due to their specific adaptation local farmers depend on for their livelihoods. The more commercial the sector is the greater the threat of loss of biodiversity. Niche market production through local breeds serving specific local consumer preferences may control these negative effects. Careful production system analysis is therefore demanded to conclude on the impact of gene flow on biodiversity. As mentioned above, apart from the main stream of gene flow of high-yielding genotypes, gene flow takes place from countries, regardless if developed or developing, which control some unique genotypes. The formation of some high-yielding breeds from developed countries themselves depended on the introduction of foreign genetic material, for example the formation of European and North American commercial pig breeds which heavily depended on the introduction of genetic material from Asia. Research and commercial interests point at such genotypes of current and future importance. Global transfer of these genotypes generally contributes to biodiversity and particularly if these genotypes are threatened in their native countries. Niche market production through foreign breeds with specific attributes serving local market preferences are additional examples of benefits of biodiversity through transfer of unique genotypes. Limitations of the study The most limiting factor of the study was the information basis of current gene flow. From countries with highly organised breeding structures statistical data bases on breeding stock transfer and changes in national population composition were expected. For Europe, such statistic in terms of the Eurostat statistic database was available and accessible. However, the records allowed only limited interpretation with regard to pig gene flow. First of all, MAIN FINDINGS OF GLOBAL STUDY 47 recording was not done at breed level so that necessary interpretations on the influence of single breeds were not possible. Furthermore, only the exchange of purebred breeding stock was recorded. Purebreeding statistics are however of minor importance in pig breeding where hybrids play a vital role. Additionally, with increasing degree of commercialisation of this sector, export and import statistics are less adequate to depict gene flow but company figures would be more suitable. Due to restricted information policy of international breeding companies this information in general lacks. In commercialised breeding sectors as the case in the pig breeding, gene flow additionally underestimates the global influence of single breeds when the breeds are being multiplied in national centres. Data on the exchange of semen were in general lacking. The Country Reports on the State of Animal Genetic Resources in some cases compiled necessary information. However, if country reports included data on animal transfer, the usual focus was on imports put not on exports. Exports could only be traced indirectly as imports into other countries. Therefore, an important source of recorded information on gene flow was generally left out. This had particularly negative consequences with regard to information on gene flow into developing countries. Here records of animal movements do rarely exist and country reports, if existing, cannot contain the desirable detailed information. Export data from developed countries could principally have closed that information gap. This throws light on the lack of information of gene flow from developing countries particularly from South to South movements which would have constituted an interesting aspect of gene flow if depictable. In general it became evident that single transactions are impossible to follow up. However, particular in the commercialised pig breeding sector, the transaction of few animals may have major impacts on global gene flow. These examples become however only retrospectively evident as changes of national population composition making this indicator of gene flow an important tool in the study. Main information source were again the Country Reports which were therefore treated with special reference in the global study. The information content varied greatly between the available reports. Additionally, a major limitation of the country reports with regard to information on changes in national population composition was put forward in the pig case study. It was assumed that in the case of Vietnam in the country report the negative influence of foreign genotypes on local breeds were probably only realised if direct interference of both took place but that the true picture remained underestimated according to more updated sources if there was no direct interference. Due to the limited information basis, the study had to restrain from quantifying pig gene flow on a global scope. It was shown that pig gene flow is a very complex and dynamic process with many factors varying the possible impacts from case to case. The impact evaluation of global gene flow on economic development and biodiversity as treated above therefore at purpose concluded at a very general and qualitative basis. The study objectively analysis possible aspects of gene flow and is meant to provide scientifically founded information for decision makers in the ongoing discussion on global gene flow and its impacts. The study is certainly not suitable to finalise this discussion. Need for further action With regard to the limitations of the data bases accessed in this study, further actions are needed to improve the information status on current pig gene flow. The accessibility of 48 MAIN FINDINGS OF GLOBAL STUDY data of the breeding companies would need considerable improvement to allow satisfactory impact evaluation studies in future. For the improvement of data quality and data quantity in developing countries the focus should be set at deriving reliable data on national population composition changes which call for frequently carried out census studies at breed respectively genotype level. This approach would avoid the time consuming tracing of complex gene flow routes instead starts right at the impact evaluation level with regard to biodiversity and on to economic development. Additionally, many weaknesses of the information base could be minimised such as the different comprehension of the status of national genetic resources through different groups. Each country should subsequent to its evaluation of the national situation make use of the identified control mechanisms of gene flow to or from the country to manipulate the situation to its better. In developing countries biodiversity and poverty reduction are two main claims future animal production has to deal with. Concepts which serve both claims at the same time by preserving or even improving smallholder livelihoods in marginal areas based on their indigenous genetic resources, can therefore not be acknowledged high enough. From these concepts derive the need for the further identification and evaluation of the genetic resources in terms of genetic distance to other breeds, their genetic potential and performance in different production systems and their risk status. There also derives the need to take biodiversity impacts into account when framework conditions are evaluated. Each country has to be aware that unique genotypes are in global demand. Based on the depicted complexity of gene flow, it is concluded that once a genetic resource has been transferred in whatever quantity it is difficult to follow up the routes of gene flow. The control of gene flow should therefore be individually considered and instrumentalised in advance and not in retrospective. MAIN FINDINGS OF CASE STUDIES 49

4 MAIN FINDINGS OF CASE STUDIES

In the following chapters the main findings of the case studies are summarised. The extended versions of the case studies are found in the Annex chapters 9.6 to 9.9. The case studies facilitated the examination of specific aspects of gene flow. In the case of the improves Awassi and Assaf breeds from Israel it was attempted to depict the historic development and current status of gene flow of these particular breeds, and to conclude on their global impacts. The goat case study depicts the history and global development of the Anglo Nubian breed and focuses in its impact study on the specific case of smallholder farms in Bolivia. The cattle case study is based on the depiction of the origin, global transfer and utilisation of the Boran and Tuli breeds with special reference to the issue of access and benefit sharing. The pig case study represents an impact study of exotic pig breeds in Vietnam with regard to socio-economic development and biodiversity. Besides information from databases, country reports and publications, the case studies particularly depended on information from regional expert organisations.

4.1 The worldwide gene flow of the improved Awassi and Assaf breeds of sheep from Israel T. Rummel, A. Valle Zárate and E. Gootwine The sheep case study depicts the development of the Improved Awassi and Assaf breeds in Israel and their worldwide spread. The information was gathered using documents from the Ministry of Agriculture of Israel to identify destination countries and transfer stakeholders. All importing stakeholders were contacted via e-mail and phone and informal interviews were conducted about the transfers and about the development and current state of the breed since its import. If the direct stakeholder of the transfer was not available for interviewing, key persons were identified in the country and interviewed. Where published literature was available, it was used instead of - or in addition to - personal communication. In Israel, within-breed selection in the unimproved sheep started at the beginning of the 20th century by Jewish sheep breeders, resulting in the formation of the Improved Awassi breed (Figure 3). The Improved Awassi differs from the unimproved Awassi mainly by its remarkable high milk production - about 550 vs. 70 litres per lactation, respectively. To improve its prolificacy, the Improved Awassi was crossed in the 1960s with the East Friesian Milk sheep that was imported from Germany, resulting in the formation of the "Assaf" with improved prolificacy of about 0.4 lambs born per lambing over the Improved Awassi. Today, the Assaf has nearly replaced the Improved Awassi in Israel’s intensive dairy sheep sector. Recently, the Booroola gene, a major gene coding high prolificacy, was introgressed by crossbreeding to the Improved Awassi and the Assaf resulting in two new strains, the Afec Awassi and the Afec Assaf, respectively, both with a prolificacy of about 2.0-2.4 lambs born per ewe lambing. 50 MAIN FINDINGS OF CASE STUDIES

Figure 3: Historical overview on Awassi and Assaf breeding in Israel

Unimproved 5000 BC Awassi

Hirrik ewes, 1932 AC Turkey Improved Awassi East Friesian Milk 1955 AC Sheep, Germany Improved rams Assaf

Booroola Merino, New Zealand 1986 AC

Unimproved Improved Afec Assaf Afec 2004 AC Awassi Awassi Awassi Assaf

The gene flow of the Improved Awassi breed of sheep from Israel started in 1965 (Figure 4). Since then, 28 transfers to 15 different countries are documented, with summarised monetary transfers over $2 million. Totally, 5,433 lambs, 1,100 doses of frozen semen and 143 embryos of the Improved Awassi have been exported from Israel. While only 9 of these 28 transfers have been part of development aid projects, the majority represent commercial transfers between private sheep farmers or government institutions and Kibutzim in Israel. Even though the biggest part of Improved Awassi breeding material, 2,944 lambs and 1,000 doses of semen, was transferred to Eastern Europe and Central Asia (Bulgaria, Former Yugoslavia, Hungary, Romania and Kazakhstan), the highest numbers of pure Improved Awassi and crossbreds with indigenous sheep are today in Spain (150,000-200,000 sheep) and Western Australia (100,000 sheep). The transfers to the tropical countries Burma, Ethiopia and India were part of development projects, but with limited success. In the Middle East, a total of 1,113 Improved Awassi lambs have been exported to Jordan, Iran, Abu Dhabi and to Turkey, where most of them were used to improve local stocks of unimproved Awassi. The following unofficial transfers to secondary destinations cannot be traced accurately. The gene flow of the Assaf started in 1977 (Figure 4). 687 lambs, 11,354 doses of semen and 260 embryos, for a total price of $333,040, have been exported from Israel in 10 transfers to 7 different countries. Only 2 transfers were part of development aid, while the majority were commercial imports. The main stream of the gene flow of the Assaf breed of sheep went to the Iberian Peninsular. Portugal imported 6,854 doses of frozen semen and 260 embryos in the years 1991/1992, while Spain imported 430 lambs and 4,000 doses of semen between 1977 and 1993. In these countries the Assaf is today the dominant dairy sheep breed, numbering over 1.2 million pure- and crossbreds. Besides the two Iberian countries, the Assaf has been transferred to Peru, Jordan and Abu Dhabi, but with much less effect on the sheep production sector. From Portugal, Assaf breeding material was exported to the United Kingdom and Italy. No Assaf breeding material has been transferred to Eastern Europe, Central Asia, Australia or New Zealand. MAIN FINDINGS OF CASE STUDIES 51

Figure 4: World wide gene flow of the Improved Awassi and Assaf breeds of sheep from Israel

United Kingdom Albania Spain Hungary Mediterranean East Europe and Countries Bulgaria Central Asia Portugal Romania Kyrgyzstan Former Yugoslavia Italy Kazakhstan Cyprus Turkey Iran

Israel Middle East

Jordan Abu Dabi Peru

Ethiopia

India New Tropical Zealand Burma Countries

Australia Okinawa

From the history and development of the Awassi of Israel, the following conclusions can be drawn: The Improved Awassi in Israel is a successful example of within-breed selection, without losing the adaptation of a breed to its natural environment. It also shows the genetic potential which can slumber in indigenous breeds. Developing the Assaf breed through crossbreeding with imported East Frisian Milk Sheep gives an example of a well adapted composite breed. The Afec Awassi and Afec Assaf lines bred through introducing the Booroola gene from New Zealand represent a very specific gene flow where only a major gene was transferred to improve a specific trait: prolificacy. This kind of single gene flow may dominate future breeding activities. Both the Assaf breed and the Afec lines show the impact of economic pressure on sheep breeding. The social structure of the agricultural sector in Israel, dominated by co-operative farm structures, had a decisive impact on the formation of the Improved Awassi and Assaf. The fact that developing the Improved Awassi breed and its later gene flow has begun by the return of the Jewish people to Israel gives an example of the impact of human migration on gene flow in animal genetic resources. Looking at the gene flow of the Improved Awassi and Assaf breeds of sheep from Israel during the last 50 years and the development of the breeds in the different countries of destination, the following conclusions can be drawn: In general, the gene flow of the Improved Awassi and Assaf breed of sheep is characterised by free animal movements, based on commercial interests, with a minimum of government involvement. With a few exceptions, where the transfers were part of a bilateral co-operation program, the movements of these breeds were commercial transfers with freely agreed benefits shared 52 MAIN FINDINGS OF CASE STUDIES by stakeholders, exporters and importers alike. In all cases animals were purchased. Although Israel has an interest in exporting developed breeds, most of the transfers were started by importers aiming to improve their national production systems. Development projects, emergency aid and government institutions played a negligible role in the gene flow of the Improved Awassi. The most successful developments of transferred Improved Awassi and Assaf breeds were organised by private persons, breeders or companies (Portugal, Spain, Australia, Hungary), while those organised by governmental institutions and NGOs seem to have had less sustained impact on the national sheep sectors (Ethiopia, India, Jordan, Iran). Concerning the geographical direction of the transfers, the case study on the Improved Awassi breeds describes gene flows from south to north, north to south and from south to south. In terms of numbers of animals established the emphasis was on the south to north movement. Today the majority of Improved Awassi and Assaf stocks are kept in European countries, north of Israel. Farmers of the Iberian Peninsular, where the Assaf is the dominating milk sheep, had the highest benefit from the past gene flow of Awassi and Assaf breeding material. While the south to north movement is dominated by the Assaf breed, in the movement to the South the Improved Awassi plays the major role (Australia). In the case of the Mediterranean countries, Improved Awassi and Assaf imports had a serious impact on the traditional breed structure of milk sheep. With increasing numbers of Assaf and Awassi during the past decade, the numbers of traditional milk breeds like the Serra de Estrela in Portugal and the Churra and Castellana sheep in Spain has decreased dramatically. But since those indigenous breeds are still not endangered, the Awassi and Assaf can be considered as contributions to the genetic diversity of these countries. In other places, like Australia, the transfer had no impact on the mutton breed structure at all, but the new breed exploited a specialised market (Middle Eastern premium fat-tail lamb market), to which no traditional breed fitted. The latter example shows the feedback gene flows can have on the regions from which they originate: the established Improved Awassi stock of Australia already supplies the Middle Eastern market with premium fat-tail lamb through considerable live exports to these countries, which will increase with growing stock numbers in Australia, and this competes with local sheep breeders in the Middle East. An important stimulus for the gene flow of the Improved Awassi and Assaf breeds of sheep was the development and spread of artificial insemination (AI) and embryo transfer technology, creating an efficient tool to replace expensive live animal transfers and to reduce the threat of importing foreign diseases. The rapid increase of the breed on the Iberian Peninsula would not have been possible without those technologies. Similarly, introducing the Improved Awassi via live animal import to Australia would probably have been denied by the national veterinary departments, in order to prevent foreign diseases being imported with them and threatening the national sheep production sector. An additional aspect of gene flows in animal genetic resources can be seen in the transfers of Improved Awassi and Assaf breeding material to Jordan and to the Palestinian Territories. In addition to their economic benefits for the local sheep breeders, those transfers have been part of an ongoing peace process and international co-operation. Sharing achievements, in this case in animal production, can build confidence. But these examples also show that politically motivated actions are often less successful than commercial ones. MAIN FINDINGS OF CASE STUDIES 53

With intensive management on station, the Improved Awassi showed a higher level of performance in most countries in milk and mutton compared to indigenous breeds. But in several examples, especially in the tropical countries India and Ethiopia, but also in Eastern European countries, the Improved Awassi could not maintain its level of performance in local production systems with low input. This shows the importance of production system analysis before introducing a foreign breed or crossbred, in order to prevent a mismatch of breed and production system. Gene flow of animal genetic resources often includes a flow of animal diseases. Local breeds are often not adapted to exotic diseases. In developing countries, where the need to raise the performance of local breeds in order to provide food is high, import regulations and restrictions have often been insufficient, as was seen in Ethiopia. On the other hand, government restrictions and import policies can be major obstacles to gene flow, as seen in Portugal and Australia. In Portugal, exporting Assaf breeding material is increasingly difficult due to tightening European laws for genetic transfers. In the case of Australia, government restrictions and veterinary requirements increased the organisational and monetary investments dramatically, restricting animal transfers only to people or organisations with sufficiently high financial capital. Examples for a lack of fit to the production environment are the attempts of introducing the Improved Awassi to India and Romania, while the cases of Ethiopia and Britain illustrate a lack of fit between breed and production system or environment/climate conditions.

4.2 History and worldwide development of Anglo Nubian goats and their impacts in smallholder farms in Bolivia A. Stemmer, C. Gall, A. Valle Zárate The Anglo Nubian is an example of a breed developed by combining genetic resources from different parts of the world to optimise performance and adaptation to tropical conditions. It is a dual-purpose goat used for milk and meat production. Its origin can be traced to Nubian goats and the Indian dairy breed Jamnapari, which may have a common ancestor in ancient Iran. The Nubian group includes the Zaraibi, Damascus and Sudanese Nubian. Herdbooks of the Anglo Nubian are kept in Britain, the USA, Canada and Australia. The countries of the South with purebred or crossbred Anglo Nubians have no such official records and information on numbers is scarce or non-existent. Information was compiled through available project reports, literature, statistical records where available and accessible and interviews with experts. The initial crossbreeding, which led to the formation of the Anglo Nubian breed took place during the second half of the 19th century. In 1910, the Anglo Nubian was recognised as a breed in England and registry began. Anglo Nubians were exported for the first time from Britain to the USA in 1909 (Figure 5), reaching a total of about 30 goats up to 1950. Here, Anglo Nubians were bred and selected without any further crossbreeding with other breeds. In Canada, a breeding programme was established in 1921, based on imports of Anglo Nubians from Britain. Offspring were imported into the USA and continued to have a great impact on Anglo Nubians there until the late 1940s. From the USA Anglo Nubians were exported to Puerto Rico and Latin America as early as the 1940s. Later on, Anglo Nubians were exported from Britain and the USA in several development efforts in Latin America, 54 MAIN FINDINGS OF CASE STUDIES

Africa and Asia. In some countries, Anglo Nubians continue to be kept as purebreds (Mexico, Brazil, Peru, Colombia, several Caribbean states, Egypt, Israel, Oman, India, Bangladesh, The Philippines, Mauritius and Malaysia) although numbers are sometimes so small that it is difficult to conserve the population (Venezuela, Ecuador, Thailand). More widespread is the use of the Anglo Nubian in crossbreeding. In some countries, like Chile, imported Anglo Nubians together with other specialised breeds caused a decline in number of the local Criollo goat.

Figure 5: Gene flow of the Anglo Nubian goat

Canada High volume

USA Europe Low volume

Mexico

Belice North and Central Israel America Middle East Oman Asia

British Caribbean India Bangladesh Venezuela

Guayana Malasia Ecuador

Perú Brasil Australia Oceania Bolivia Chile Ethiopia New Zealand Caribbean and Argentina South America Kenya

Africa

In Bolivia, there are no official records on the introduction of goats from outside Bolivia. The case study is based on information compiled from project reports and personal telephone or e-mail interviews with development workers of governmental and non- governmental organisations, goat keepers and other experts. Goats in Bolivia are mainly kept in the inter-Andean valleys at altitudes ranging from 1,000 to 3,000 meters. Anglo Nubian goats have been imported to Bolivia in the late 1960s up to the present. Animals originated in Argentina, Brazil, Paraguay or the USA; semen was introduced from Germany. There have also been exchanges between different parts of the country. Anglo Nubians were introduced in the regions with a tradition in goat keeping, namely the departments of Cochabamba, Tarija, Chuquisaca, Potosi, and some parts of Santa Cruz. The majority of imports and exchanges within Bolivia were planned and handled by development agencies (6 cases), non-governmental organisations (3 cases) and MAIN FINDINGS OF CASE STUDIES 55 universities (2 cases) while only in a few cases did individual goat owners import Anglo Nubians. Successes or failures of these introductions of Anglo Nubians to different production systems and climates can be summarised as follows: Anglo Nubian populations increased in intensive or semi-intensive management systems including health care, improved housing and fodder production. These conditions are found in institutional flocks or those managed by private entrepreneurs. In these intensively managed goat farms, purebred Anglo Nubians have positive impacts because of their good adaptation to the climate of the inter-Andean valleys and their high yields. In semi-extensive management, this breed is able to survive but its production potential is not fully realised. There are only few examples of persistent Anglo Nubian populations in these conditions, and all of them include some access to the market so that goat owners can sell cheeses at least during part of the year. Another prerequisite is the extension service needed to assist goat owners with the introduced breed so that they can respond to the higher needs of Anglo Nubians compared to the Criollo goats in terms of better housing, parasite control and improved feeding. In extensive management, Anglo Nubians did not last for long. The time needed to adapt to a new environment can be very long when adult animals are transferred from one climate to another, in some cases only the offspring of the introduced animals adapt well. Purebreds could not be transferred to smallholding conditions without mortalities surpassing 50% and high abortion rates. When introduction is in the form of a slow process using young pure or halfbred bucks for breeding, results are much more favourable. Where animals were given to goat keepers as development aid, no direct monetary loss for the farmers was recorded. Introducing crossbreds can result in benefits to those smallholders who are able to make use of the higher milk production by marketing part of the cheeses produced. A possible negative impact is the greater need of the crossbreds for improved management, which translates into higher monetary input. The impact of introducing Anglo Nubians on the Criollo goat population has been very limited in Bolivia. Today, Criollo goats far outnumber Anglo Nubians and crossbreds. A negative impact on the Criollo goat population, though, can be seen in the efforts and financing that was spent in introducing Anglo Nubians to climates and management systems not suitable to the breed. Had the same effort been spent in their improved breeding and management, economic impact might have been higher. The impact of introducing Anglo Nubians on the environment was positive in those cases where there was a simultaneous improvement of the pastures caused by rotating flocks and building enclosures. Inputs in infrastructure and management were responded to by high performance of the improved breed, making sense of restricting flock sizes and movements of the animals. The gene flow between developed countries was realised by private persons and on a commercial basis, whereas the transfers from developed to developing countries often involved development agencies. In the case of Bolivia, the gene flow was planned mainly by agencies, and only in few instances by goat keepers. There were also gene flows between countries of South America. 56 MAIN FINDINGS OF CASE STUDIES

In Bolivia, the introduction of Anglo Nubians was met with failure when the planning was done merely by development agencies without considering or even knowing the production systems involved. On the other hand, careful planning and involvement of the goat keepers themselves led to success. Finance for introducing the Anglo Nubians came mostly from development agencies, governmental or non-governmental. Only in few cases did private financing occur. It is concluded that in most cases, smallholder goat keepers did not suffer financial losses due to the introduction of Anglo Nubians, however public incentives frequently did not pay back. The negative aspect of the transfer of Anglo Nubians lies in the effort spent introducing and adapting a foreign breed to the conditions of Bolivia, while the local genetic resource, the Criollo goat, has been neglected both by researchers and development agencies. The Criollo has a high variability and production potential that can be utilised if these animals are included in research on management and breeding improvement.

4.3 Boran and Tuli cattle breeds - origin, worldwide transfer, utilisation and the issue of Access and Benefit Sharing S. Homann, J.H. Maritz, C.G. Hülsebusch, K. Meyn, A. Valle Zárate African indigenous cattle breeds, particularly the Boran and the Tuli, have received increasing interest in the past as a source of genetic diversity with potential to improve cattle production in sub-/tropical environments worldwide. In this context, controversy has arisen about conservation-through-utilisation strategies and access and benefit sharing. The present case study attempts to describe the worldwide transfer of genetic material of the Boran from East Africa and the Tuli from Southern Africa. The so-called Improved Boran is the result of intensive ongoing breeding started by British settlers in Kenya from the indigenous Boran cattle stock of southern Ethiopia, south-eastern Somalia and northern Kenya. The Tuli evolved from indigenous Sanga cattle from Zimbabwe and eastern Botswana, and was maintained and improved on state breeding and research stations in Zimbabwe. Available information was gathered from the scientific literature and from the internet. Networks involved in using and conserving cattle genetic resources were identified and semi-structured interviews were held with key people involved in the breeding of, research into, and utilisation and conservation of the Boran and Tuli. The aim of this study was to give an overview of the origin and movements of the two breeds (Figure 6 and Figure 7). Their worldwide distribution, performance and further development in other countries are reviewed. The formation of Boran and Tuli cattle breeders’ societies and their national and international networks are reviewed, as well as government breeding and trade policies and their impact on the further development of the breeds. Data on imported and exported germplasm as well as the current livestock populations are quantified for each country where possible. Issues concerning the modes of transfer of genetic material and access and benefit sharing for livestock genetic resources are addressed. MAIN FINDINGS OF CASE STUDIES 57

Figure 6: Worldwide transfers of Unimproved and Improved Boran cattle breeding material from Eastern and Southern Africa

USA

Nigeria Ethiopia Mexico Somalia

Brazil Uganda Kenya Eastern Africa North- and South America Congo Tanzania

Zambia Zimbabwe

South Africa Swaziland

Unimproved Boran Australia

Improved Boran Australia Southern Africa

Figure 7: Worldwide transfers of Tuli cattle breeding material from Southern Africa

North- and South America Canada

USA

Mexico Gabon Australia Argentina Australia

Namibia Botswana Zimbabwe

Southern Africa South Africa Tuli

The Boran (Bos indicus) is a major cattle breed in eastern Africa, originating in the Borana rangelands in southern Ethiopia. The Tuli cattle (Bos taurus) descend from a small nucleus herd of yellow Ngwato cattle in Zimbabwe. Both breeds have evolved under harsh arid and semi-arid range conditions, and out of the local herders’ indigenous breeding and selection strategies. In Kenya, subsequent breeding of the Boran cattle was started by British settlers, leading to the formation of the Improved Boran. In Zimbabwe, the Tuli were further developed at government breeding and research stations from the early 20th century. 58 MAIN FINDINGS OF CASE STUDIES

Breeding organisations were founded in the regions of origin, and research and breeding programmes were set up to improve beef production while maintaining adaptability to environmental constraints. Boran cattle showed high fertility and production in low-input environments, while Tuli cattle also showed high fertility and excellent beef quality, but produced comparatively better in high-input environments. Because of their adaptability and productivity in tropical conditions, Boran and Tuli cattle attracted the interest of livestock scientists and the international beef industry. In 1988, in order to add breeding options to composite crossbreed populations for industrial beef production, Australian researchers - collaborating with a consortium of Australian beef producers - imported the first Boran and Tuli embryos from Zambia and Zimbabwe. In 1991, Boran and Tuli embryos were exported from Australia to research stations in Nebraska and Texas, USA, where the largest germplasm comparison programmes on beef cattle in the world were undertaken. The breeds also found their way into the Australian, American, and South American beef industries through various other channels. The majority of documented transfers of Boran and Tuli genetic material represent commercial transfers between government research stations and private enterprises and among international business partners. The Tuli appears to have been better accepted among beef producers in Australia and America than the Boran. However, comprehensive data about the current population and the contribution of both breeds in crossbreeding schemes or in the formation of composite breeds are scarce. From the few available figures, the population of these breeds in these countries is small. Given the overall size of the beef industry, with 94.9 million head of beef cattle in the USA and 26.4 million head in Australia in 2004, the likely contribution of Boran and Tuli cattle to the sector appears insignificant. However, the study shows that a demand exists for genetic material with adaptive traits and special traits of beef quality. In Africa, Boran and Tuli cattle were used systematically to a considerable extent for commercial ranch development, but not for local or regional livestock improvement schemes. Today, Boran and Tuli cattle are said to face genetic dilution in their areas of origin, although they are still used - particularly the Improved Boran in Kenya - for commercial beef production. Some Non-Governmental Organisations argue that the Boran and Tuli were exploited by networks of institutions, business companies and individuals, without sufficient compensation to the original breeders. They have raised a controversial debate on access and benefit sharing agreements, which is dominated by three main issues: Whether or not a prior informed consent existed before the initial transfers from Africa into Australia; whether or not the call for additional compensatory payments to the original breeders is justified; and how big the impact of Boran and Tuli genetic material on the Australian and North American beef sector actually was. The first two issues are ethical rather than scientific, and have to be politically resolved. The actual contribution of the Boran and Tuli to upgrading the Australian and American beef sector appears to be negligible. It did not nearly meet the high expectations which triggered the operation, and the actual use of Boran and Tuli cattle in the Australian and American beef sector is limited to singular cases. Prior informed consent should be more carefully looked for in future transactions to avoid a posteriori claims. MAIN FINDINGS OF CASE STUDIES 59

From the history and development of the Boran and Tuli cattle breeds, the following conclusions to be drawn: Improving the Boran and Tuli cattle breeds in their African environment is an example of successful within-breed selection and dissemination, maintaining a strong resistance to environmental stress. It is also an example for the comparative advantage of locally adapted breeds over imported exotic breeds. The genetic material from the Boran and Tuli breeds - despite high expectations - does not seem to have significantly contributed to upgrading the international beef industry, although the actual impact cannot be quantified with the available data. However, the use of these breeds is an example for international gene transfers and the existing demand of industrial beef producers for new genetic material, which has particular beef quality traits and traits of adaptation to harsh environments. The trade in Boran and Tuli genetic material has raised awareness of the potential of germplasm from developing countries for livestock industry sectors in developed countries, and underlines the advantage of conserving livestock biodiversity for the future. The example has drawn public attention back to the value of Boran and Tuli purebreds in Africa. It has also contributed to raise awareness for adaptive traits and for beef quality traits in cattle breeding. Embryo transfer and artificial insemination were the prerequisite for international trade of Boran and Tuli genetic material in that these techniques have overcome the sanitary barriers between the African countries and Australia. They have also reduced the costs and burden of live animal exports. This was crucial for the initial dispersion of Boran and Tuli germplasm to the world market. Vertical integration of multiple stakeholders - research institutions, government agencies, beef industry and individual business people - has essentially contributed to the initial transfer of Boran and Tuli genetic material to industrial countries as well as to their further marketing, lobbying and evaluation. Vertical integration could likewise be applied to breed conservation initiatives, involving local livestock keepers, breeders’ societies, universities, governments and development agencies. The transfers of Boran and Tuli genetic material are characterised by free movements with governmental impact limited to livestock sanitary / veterinary regulations. Most transfers are between commercial stakeholders, and freely agreed benefit sharing can be assumed. Geographically the main streams were between Australia and America, and South Africa as well as within Africa. The countries of origin of the Boran and Tuli breeds are not the major players in this, however limited, business. The issues on whether or not a prior informed consent existed before the initial transfers of Boran and Tuli genetic material from Africa into Australia, and whether or not the call for additional compensations to the original breeders is justified are controversial discussions. The main positions raised in the debate over the Boran and Tuli case are speculative and not based on sound scientific investigations. These issues are ethical, and have to be resolved politically. The actual contribution of the Boran and Tuli to upgrading the Australian and American beef sector appears to be negligible. It did not nearly meet the high expectations which triggered the operation and the actual use of Boran and Tuli cattle in the Australian and 60 MAIN FINDINGS OF CASE STUDIES

American beef sectors is limited to singular cases. Prior informed consent should be more carefully looked for in future transactions to avoid a posteriori claims.

4.4 Impact of the use of exotic compared to local pig breeds on socio-economic development and biodiversity in Vietnam Le Thi Thanh Huyen, Regina Roessler, Ute Lemke, Anne Valle Zárate This case study focuses on the distribution of the main indigenous pig breeds and crossbreds in Vietnam, the introduction of high performance breeds and their impact on biodiversity, and the suitability of different breeds for different environments. Vietnam owns a wide variety of local pig breeds across different regions of the country. The Lang Hong, Mong Cai and I breed are the product of a long deliberate breeding history, whereas other breeds, e.g. the Meo, Co, or Soc, were not systematically bred. In particular the I and later the Mong Cai were strongly promoted in Vietnam to replace lower yielding local breeds. In South Vietnam, the Thuoc Nhieu, Ba Xuyen and Phu Khanh composite breeds developed from crossbreeding local with exotic pigs. The DBI-81 and BSI-81 were developed in North Vietnam from crossbreeding I sows with exotic boars, but did not become widespread in national pig production. Only the Mong Cai has become common, being now the major local sow line in Vietnam. Exotic pigs, including Large White, , Duroc and Berkshire, have been introduced to Vietnam from American and European countries since before the 1920s. Major driving forces were the French Colonial Rulers (before 1954), American forces (before 1973), the socialist government (since 1954), Vietnamese and foreign commercial companies (before 1954 and after 1986), and developmental projects (after 1986). Gene flow now and recently is mainly a net- inflow of exotic pigs. Current development and poverty alleviation projects at village level usually promote exotics, and only occasionally improved Vietnamese breeds. Information on pig gene flow to and within Vietnam is limited, due to the restricted information policy of both international breeding companies and Vietnamese official sources, but also due to the decentralised nature of pig breed import and distribution. At present, exotic and crossbred pigs dominate, while local pigs make up still 26% of the national pig herd, mostly in uplands, rural and remote areas. The decentralised structure of the Vietnamese breeding system, the less developed central coordination and the common use of artificial insemination have all supported the spread of exotic pigs in Vietnam, especially at the smallholder level, which makes up 80 to 95% of Vietnamese pig production. Smallholder pig production includes different intensity levels. In contrast with large-scale commercial pig production, they can be characterised as low-input systems. Local pigs yield lower reproductive and growth performances. Performance data in literature are rarely comparable, as local breeds have usually been investigated in low-input extensive farming conditions, while exotic pigs or crossbreds are mostly tested under improved conditions or on station. Mong Cai sows under smallholder conditions yield higher reproductive performances than exotic or crossbred pigs, implying better reproductive performance potential of local breeds. Additionally, favourable adaptation traits (regarding environmental/climatic factors, low-input production conditions, and susceptibility to disease) and general robustness are described for local pig breeds, together with favourable meat quality traits. Other, less favourable traits of local pig breeds include a high fat content and low lean meat ratio, a low growth rate, and, apart from the Mong Cai, a low MAIN FINDINGS OF CASE STUDIES 61 fertility, rendering them less suitable to respond to higher inputs, unless their special quality traits are rewarded by the consumer. The introduction of pigs and breeds from neighbouring countries (Laos, Cambodia, China) started probably centuries ago, as part of human migration (e.g. Thai and H’mong migrating from China), occupation (China), and trade. The influx of breeds was an important component in developing Vietnamese local breeds. However, information is lacking on those early phases. The earliest confirmed information on introducing pig breeds goes back to the 1920s. Gene flow in the recent past and present has probably been a net-inflow of pigs. Exports (e.g. Vietnamese potbellied pigs to western countries as pets and for scientific use) were negligible. Before 1955 (end of French colonisation) and after 1986 (economic liberalisation), pig imports were directed by commercial interests as the main driving force of gene flow. From 1955 until 1986 the major driving force was the policies of the socialist government, and after 1990 additionally foreign developmental projects, both with the declared aim to benefit the poor farmers, but not always fulfilling their claim. The inflow of pig breeds to Vietnam consisted of higher-yielding breeds from Europe and America, which were introduced due to their higher performances (in the countries of origin) to improve or replace the low yielding local breeds. Commercial imports consisted of exotic pigs. Current development and poverty alleviation projects at village level usually promote exotics, and only occasionally improved Vietnamese breeds. Information on pig gene flow to and within Vietnam is limited, due to the restricted information policy of both international breeding companies and Vietnamese official sources, but also due to the decentralised nature of pig breed import and distribution. The introduction of exotic pigs was supported by the decentralised nature of the Vietnamese breeding system. Centralised coordination of breeding measures is not well developed, and centralised measures fulfilled their aims only partly. However, the impact of the state-run breeding stations has been considerable; and the advanced use of AI has strongly supported the introduction of exotic genetics to the smallholder producer. The influx of exotic breeds has positively influenced output and efficiency of pork production in Vietnam, while the local pig populations have been reduced. Today, pigs of various crossbreeding degrees are widely distributed. Most indigenous breeds show declining population trends, and the majority of local breeds are in a vulnerable or critical condition or even face extinction. Conservation measures of Vietnamese institutions follow suitable approaches (in-situ conservation on-farm). However, due to shortcomings in set- up and implementation, they may not successfully preserve local pig breeds. The long-term sustainability of those programs is questionable. National decisions and the willingness to pay for conservation programs depend on expectations for future benefits, which need to be based on scientific proofs of the value of specific traits, and market-backed valuations of products. The significant genetic distinctions both between Vietnamese breeds and between Vietnamese and European breeds have been shown. Local breeds are a source of promising alleles, which might be significant for future genetic improvement and of unpredictable economic value. 62 MAIN FINDINGS OF CASE STUDIES

Local pig breeds are a significant component of the Vietnamese and worldwide biodiversity, and are still important for resource-poor farmers in Vietnam, who depend on them to ensure their livelihoods. The dominance of high-yielding exotic breeds will increase in intensified production systems. Local breeds will only contribute to worldwide biodiversity if their competitiveness to exotics is proved for production systems under development and/or if favourable adaptation traits are proved and the controlling alleles identified. Investigations are under way to define local pig breeds, characterise them, and compare their performances under standardised conditions. Research results indicate a considerable production potential of local pig breeds especially under low-input conditions, favourable adaptation traits, and genetic peculiarities, differentiating them from the European breeds. Local pig breeds are a significant component of the Vietnamese and worldwide biodiversity, are important for resource-poor farmers in Vietnam who depend on local breeds to ensure their livelihoods, and for future breeding measures utilising e.g. favourable adaptation traits. On the other hand, exotic pigs have become increasingly available and accessible to farmers in Vietnam and have enabled them to produce pork with increasing efficiency. Whether pig-keeping resource-poor smallholders in remote and mountainous regions can be integrated, or if they can set up niche production with local pig breeds, remains to be clarified by further investigations. Further investigations are required to define local pig breeds, further characterise their genetic specificities, and to comparatively evaluate their performances under standardised conditions.

ANALYSIS OF DRIVING FACTORS AND IMPACTS 63

5 ANALYSIS OF DRIVING FACTORS AND IMPACTS

5.1 Main factors affecting gene flow Analysing global studies and case studies, we identified the following main factors that affect gene flow. As gene flows of the four species are affected differently by these factors, subchapters refer to them separately in the order sheep, goats, cattle and pigs.

5.1.1 Breeding and trade organisations

Breeding and trade organisations impact gene flow with breeding methods, sales strategies, and in particular distribution channels and their competition. For breeding, the genetic quality, targeted quantity and diversity of the breeding products are important factors. Suitable breeding programmes require central breeding management and laboratories to use biotechnology, which require strongly developed breeding associations, commercial breeding enterprises or government breeding institutions. Regional breeding organisations, aiming to be self-sufficient in a regionally closed system, are vanishing because they cannot compete. Concerning distribution channels, indirect channels are those where independent trade organisations operate between breeding organisation and customer, locally or internationally. Direct channels are those where the breeding organisations distribute their products themselves to national and international markets. Directly controlling end sales by the breeding organisation is particularly attractive to commercially aware organisations. Approaches to international distribution include franchises, joint ventures and subsidiaries. Increased sale volumes, and so increased breeding populations, leads to strategic alliances and enterprise networks. Breeding and trade organisations tend to concentrate in order to reduce costs, further increase genetic gain, and secure sale volumes. This concentration can become critical if enterprises become large enough to control markets and competitive access to those markets is difficult.

5.1.1.1 Sheep

Similarly, sheep breeding organisation influences the development of superior genetic material through many factors. The proportional share of the world’s total number of sheep breeds in each region as published by the FAO in the World Watch List for Domestic Animal Diversity shows that 47.9% of the world’s sheep breeds are found in Europe (Annex 9.1 A 54). This reflects the advanced breeding organisations of European countries, and indicates why developed countries supply most introduced breeds. The increasing formation of composite breeds after the 1960s studied by Shrestha (2005) (Annex 9.1 A 4) also revealed that the foundation breeds used were almost all developed in the western world. The material available on current gene flow, compiled in chapter 3, indicates that there is an influx of genetic material into many regions of the world. Almost all transferred breeding material was reported to belong to breeds developed in countries with advanced breeding structures. This origin was not only true for gene flow going to countries with temperate production environments but also genetic material going to the tropical and 64 ANALYSIS OF DRIVING FACTORS AND IMPACTS subtropical region. Examples were the Improved Awassi and the Dorper breed developed in Israel and South Africa for the subtropical and tropical production environment. Access to improved breeding material was however not limited to developed countries; destination countries were as much developing as developed. The example of the Improved Awassi and Assaf breed given in the case study shows that the direction of gene flow transfer was from south to north, north to south and south to south. However, the majority of Improved Awassi and Assaf stocks are now kept in European countries. Farmers of the Iberian Peninsular had the most benefit from the gene flow, indicating that breeding organisation may not limit the direction but does influence the impact. The breeding organisation has obviously a significant influence on gene flow. The more advanced the breeding structure of a country is, the greater the chance to develop and spread superior breeding material, adequate trading organisation assumed. Therefore, developed countries have an advantage over developing countries in developing and spreading improved breeding stock. Australia’s Merino breeding organisation and its dominant role in supplying superior Merino breeding material for many countries world wide is the most striking example in the sheep sector. The Argentinean Merino for example is genetically almost identical to the Australian Merino (Mueller, 2004). However, gene flow from Argentinean Merino breeding stock is limited to some exchange with other Mercosur member countries. Access to improved genetic material however does not depend on breeding organisation. Both developed and developing countries are in demand for new breeds. Organising breeding in developing countries will therefore not reduce demand for foreign breeds, but on the contrary increases it. This is regarded as a basic prerequisite to gain shares in the global market of genetic resources. Gene flow from developing countries was observed when a superior genetic resource was available without preceded systematic improvements. The export of the D’man sheep from Morocco is an example. A superior genetic resource may be available in any country without preceded systematic improvements. This has been observed in breeds with superior prolificacy and is often based on the action of major genes. If a nation intends to exploit this genetic resource, it has to be aware of the competition in the global genetic material market. The gene flow of the Booroola gene is one example. Despite the fact that Australia had placed a national moratorium, the genetic resource continued to be exploited by other countries that already had it at their disposal. Similar acting genes for prolificacy exist in other local populations in southern as well as northern countries, but are limited in their distribution probably due to the lack of adequate breeding and trading organisations for this local genetic resource. Since the impact of trading organisation was not studied in detail for the sheep sector, conclusions are limited. However, the Awassi case study showed that the imported breed population was mainly affected by breeding organisation and private interests in the genetic resource, relatively independent of the quantity of transferred breeding stock.

5.1.1.2 Goats

After successfully introducing Swiss breeds to southern Germany, the first breeder associations were formed at the end of the 19th century. The same development can be observed in many other European countries. Over the years, changes in economies have stimulated changes to goat populations by breeders and breeding organisations. While ANALYSIS OF DRIVING FACTORS AND IMPACTS 65 breeding activities during the 20th century tended to consolidate if not unify goat breeds, a countermovement can be seen today: rare, peculiar breeds and some with very special characteristics are transferred from their home areas and bred for pleasure or production. Examples are the African Dwarf breed in several European countries or the and Ovambo in Germany, which are imported by hobby breeders. In New Zealand the Goat Council was created in 1979 with the aim to promote awareness of the potential of goat farming, particularly fibre production. The number of farmed goats increased from an estimated 20,000 in 1980 to 2 million by 1990 (Porter, 1996). Due to the decreasing prices of mohair and the increased demand for goat meat on the international market, the Council encouraged livestock farmers to increase the national goat herd and goat meat production for export. South African Boer goats have been imported since 1993 for its conformation, early maturity and superior growth rates. However, according to FAO statistics, in the 1990s goat numbers declined to 228,000 in 1998, reaching 150,000 in 2002; goat meat production went down from a high of about 3,000 tons in 1991 to 1,300 in 2004. Breeders’ organisations that play an active role in goat genetic improvements are not present in all countries due to socio-economic situations. In countries without breeders’ organisations, individual breeders may import breeding stock but they are often restricted by regulations on stock movement (tariff and non-tariff, organisation, cost etc.). Private firms specialising in (international) livestock trade may be of great help, but goats lack the critical mass to attract international firms to engage in their trade.

5.1.1.3 Cattle

As stated in the global study, the current development and trade of cattle semen - particularly dairy cattle semen - is a new and very intensive gene flow similar to that in and pigs. This is fostered through the high level of breeding and trade organisation, whether carried out by breeding organisations as in European countries, commercial breeding enterprises as in North America, or government breeding institutions or related organisations as in socialist or former socialist countries. The headquarters of internationally competitive and active organisations are almost exclusively located in industrial countries, which gives them an advantage over developing countries to exploit and disseminate their genetic resources. At the same time, cattle breeding and trade organisations have concentrated remarkably in the last decades. Narrod and Fuglie (2000) report that the USA has only a few companies with livestock breeding programmes. The number of companies providing AI declined from about 200 in 1950 to approximately 20 in the 1980s. In dairy and beef cattle there were only 6 top suppliers of AI in 1996. Several breeding companies are subsidiaries of other companies, and some are USA affiliates of European or Asian firms. Some of these companies merged with earlier breeding companies and the original trademark names of breeds may still be used for brand recognition. These organisations are increasingly integrated by contracts, particularly in genetic material exchange. Many indicators show that animal breeding, and breeding product trade, continues to concentrate and intensify. The result is high quality breeding products, widely diverse live animal products, and cryo- conserved semen and embryos, all available in large numbers. Ideal conditions to heavily impact the gene flow of a single breed across the globe. In recent years breeding 66 ANALYSIS OF DRIVING FACTORS AND IMPACTS companies have extended their reach from developed to developing countries. Commercial crossbred products are provided to national breeding companies often close to urban centres. Local genotypes are still provided by small-scale farmers. In general, in developing countries the classic pyramid breeding structure with a wide production sector base and the breeding sector on top is rarely established. Ideally, livestock keepers of the production sector obtain improved genetic resources from the breeding sector. They use the animals and produce breeding and production records which are the data base on which the breeding sector relies to improve breeds. Because holdings are small, lack of inputs (to realise the genetic potential of the animals) and lack of infrastructure, this system does not usually work in developing countries. In addition, without a livestock market paying premium prices for breeding animals (compared to slaughter animals) there is no incentive for breeders to invest in improving sires and dams to sell to the production sector. In this situation, breed improvement takes place only in large estates and institution herds, and gene flow to the smallholder production sector is small. Differences among countries are however large. The lower AI coverage in developing countries, due to little breeding organisation, limits access to exogenous genetic resources, again varying considerably between countries and regions. Where AI coverage in developing countries increases, gene flow of temperate breeds generally increases in proportion. However, developing breeding organisations in developing countries can also help to exploit local genetic resources. An example is the cattle breeding associations formed in East Africa as given in the case study of Boran and Tuli. Here it is shown that breeding associations can be a stepping stone to improve and develop a breed. Whether it is globally marketed and distributed depends, however, on the professionalism of the breeding, marketing, and political-economic power. Particularly national subsidies have a major influence on the animal transfer realised.

5.1.1.4 Pigs

In pig breeding, the impact of changes in breeding practices and trade organisation on gene flow is most evident. Drastic changes in the international pig market led to new breeding and sales measures. “New” production areas in Latin America, Southeast Asia, and Eastern Europe and the rapid growth of the North American agro-industry combined with trade liberalisation to increase competition and commercialisation. Pig breeding became increasingly international and concentrated. Competition required increased quality, diversity and quantity of genetic products. This called for corresponding breeding programmes and biotechnology. Both the increased professionalism and sale volumes fostered concentration further. Under the economic pressure of reduced profit margins, and to secure breeding progress and sale volumes, the concentration level rose. Countries holding leading positions, which had emerged early from the old, conservative breed-society and government-dominated structures, saw increasing monopolisation, and this gave Europe and North America considerable advantages over the rest of the world. However, nowadays there are strong breeding enterprises elsewhere, such as Thailand, the Philippines and China. Consequently, national pig breeding programmes as well as pig breeding companies have increasingly similar structures. Nucleus flocks are being installed which are limited in number but increasing in size and are integrating. Driving forces ANALYSIS OF DRIVING FACTORS AND IMPACTS 67 behind this include reducing disease risk (Foot-and-Mouth disease, classic and African swine fever are the most prominent) and diminishing exports when stricter trade regulations cannot be complied with (Brascamp, 1998). The Sanitary and Phytosanitary Measures agreement of the WTO and the OIE play a prominent role in this. With regard to gene flow, the number of breeds used in commercial pig breeding programmes has reduced to a few world wide, and their genetic variation is reduced considerably causing genetic loss. Increased concentration in breeding and trade has increased the flow of these breeds into many parts of the world. Hybrid breeding gives control of gene flow to commercial companies by maintaining their purebreeding lines and trading crossbred breeding stock which they exclude from further breeding. As pig imports and distribution are decentralised, breeding is not well coordinated. Emerging economies and developing countries can connect to international production networks - in pork following the example set by the poultry industry. As in general pig gene flow takes place in commercial production, traditional breeds are tending to be replaced through competing commercial and traditional pig production in national markets. In Vietnam for example the impact of state-run breeding stations has been considerable and the advanced use of AI has strongly supported the introduction of exotic breeds. At the same time, a national subsidy system supported the spread of exotics and subsequently the replacement of local breeds (Drucker et al. 2005). Considering socio-economic changes, imported breeding stock and breeding structures are increasingly available to farmers and enable them to produce pork with increasing efficiency. The question of whether resource-poor smallholders can be integrated, or if niche production with local pig breeds can support their livelihoods, is yet to be answered. In other regions, for example in Africa, the influence of international gene flow is very limited. In South America, the impact of imported breeds shows great regional differences, while in China rapid and unpredictable changes do not yet allow conclusions on the impact of gene flow. In developed countries on the other hand, we can see a reorientation towards local breeds. Generally, we can conclude that breeding organisation in a country - private or government - significantly determines the demand for gene flow into that country. Examples are Thailand and Brazil with private breeding organisations, and Vietnam and China with government organisations and joint ventures. On the other hand, in the pig sector, consumer preferences play an important role in the demand for breeds. Unique germplasm from Asian breeds is vital for commercial pig breeding companies from Europe and North America. Major international breeding companies need control over unique genotypes to ensure their market position. With increasing commercialisation, pig gene flow becomes harder to trace through export and import figures as these do not reflect the influence of exotic breeds. Here, company records would truly depict the influence of small trade actions that greatly impact diversity. For example, transferring a few purebreeding nucleus animals is negligible in absolute numbers but can have a considerable impact in the country of destination. 68 ANALYSIS OF DRIVING FACTORS AND IMPACTS

5.1.2 Regulations

In the 19th and 20th century, in most Central European countries governments started regulating livestock breeding and thus gene flow through laws and regulations. In order to protect the small livestock keeper who depended on commercially provided sires, livestock were subject to licensing. These measures gave breed development a strong push in the early stages. Later however, the rigid dominance of government officers proved a stumbling block to improving breeds. Similarly, the breeding organisations tended to dominate the individual breeder by setting rigid breeding goals. It is noteworthy that in the 1950s, dairy cattle breeders turned for improved genetic material to countries where governments did not interfere in livestock breeding (Gall, 2005). Today, a number of instruments are available for national and/or supranational institutions to regulate breeding and trade. These instruments include direct interventions and support, prohibitions, orders, motivation, information, consultation and education. Theoretically, regulations may influence gene flow differently. Positive external effects can be enhanced by governments directly supporting breeding organisations. Supported breeding organisations are able to increase their participation in international trade and so gene flow is stimulated. Patent protection, seen in , could be an alternative to governmental support. Negative external effects can be reduced by prohibiting breeding organisations other than particular enterprises and associations. This restricts gene flow and supports breeding and maintaining regionally adapted breeds for purebreeding. Furthermore this is an essential promotion of domestic farmer and breeding organisations. The lack of market transparency is compensated by neutral and reliable consumer information, which in turn vitalises and expands the international trade of breeding products and gene flow. Veterinary and welfare regulations play a significant role in restricting or enhancing gene flow.

5.1.2.1 Sheep

The national hygiene standards of a country’s sheep breeding sector significantly affects gene flow, via trade and breeding regulations imposed to battle transmission of transboundary animal diseases. High hygiene standards permit an almost unlimited export to all parts of the world while it prevents the introduction of stock of a lower hygienic status. Australia enjoys the rare distinction of being disease free, and so can export to virtually all parts of the world. Governments have enforced stringent quarantine standards which have prevented the introduction foreign genetic material, as it is vital for Australia to protect this highly desirable status (Padbury, 2002). AI and embryo transfers are efficient alternatives to live animal transfers that are not subject to the same restrictions. Examples include introducing the Improved Awassi, and probably the Dorper germplasm, to Australia. Argentina, as another example, applies very strict health protocols on germplasm trade. The health requirements basically restrict imports from the two countries, Australia and New Zealand (Mueller, 2004). ANALYSIS OF DRIVING FACTORS AND IMPACTS 69

In Uganda an Animal Breeding Act has been in place since December 2002. Under this act all imports and exports have to be licensed by the Ministry of Agriculture. However, illegal transfers, especially with neighbouring countries, may even have increased because of these regulations. Main diseases that restrict the current movements of live sheep are FMD, PPR and recently Scrapie.

5.1.2.2 Goats

In order to prevent the introduction of major contagious diseases governments tend to restrict imports of livestock. This applies in particular to movement between continents. Importing may be completely prohibited or subject to quarantine measures. Because of the time and cost involved stock movement, both import and export, may be effectively impossible. These restrictions give countries which are free of certain diseases an advantage for exports but may exclude affected countries as sources for imports. However, clean exporters may also be affected if they are held responsible for hygiene problems which cannot be traced back to their true source, as has been observed for small ruminants from Ethiopia. European countries encountered great difficulties importing goats from African countries where diseases are difficult to control. As the animal health situation is always changing so do the regulations on movement and import of livestock. The trend is towards severer restrictions. Australia is free of some of the major diseases, so imports are prohibited from other continents, such as those with Foot-and-Mouth disease, Transmissible Spongiform Encephalopathies or Bluetongue. Therefore, while Australian breeders are well positioned to export stud livestock, semen and embryos, they suffer import restrictions. In 1959 quarantine laws banned all goat imports except from New Zealand (Porter, 1996). Exemptions were granted only for animals passing through quarantine. However, the rigid measures, the time involved and the risk that animals may be rejected, make this is an expensive exercise. Boer goats had to spend up to 6 years in quarantine stations either on offshore islands or on mainland South Australia in the late 1980s, because ruminant virus diseases were common in Africa. While those South African goats were quarantined, some full-blooded and crossbred Boers from New Zealand were introduced instead. Quarantine requirements in Australia were relaxed in 1993-95 making it possible to import embryos from Africa (Boer Goat Breeders’ Association of Australia, 2004). Similarly, importing goats into the USA directly from South Africa was difficult until 1995 partly because of the USA-South Africa trade status but mainly because of animal health regulations. A strict five year quarantine was imposed, mainly because of Foot-and-Mouth disease. Therefore, most Angora and Boer goat imports (live animals and embryos) came from New Zealand, Australia and Canada, or are descendants of South African animals or embryos born in Canada, New Zealand, Australia and the USA (Cutrer, 1995; Machen, 1997; Smith, 2004). Europe is a major source of gene flow, mainly of dairy goats. This was severely hampered by Caprine Arthritis Encephalitis (CAE), first observed in 1980. It was eventually found in most worldwide dairy goat populations, and countries free of CAE imposed severe restrictions on goat imports. Only by applying rigid eradication measures were CAE-free populations re-established and gene flow made possible again. 70 ANALYSIS OF DRIVING FACTORS AND IMPACTS

In recent years Scrapie is increasingly a matter of concern, not only in sheep but also in goats, restricting live animal movements.

5.1.2.3 Cattle

Similarly to pigs, cattle breeding products are generally spread by a few international breeding companies and, in principle, transport of any magnitude and distance is possible. However, gene flow faces restrictions from veterinary regulations, welfare considerations for breeding animal transfers. As prominently seen for sheep, high hygiene standards of a country permits an almost unlimited trade to all parts of the world, and prevents stock imports of a lower hygienic status. For example, in the 1960s, quarantine regulations (to protect against blue-tongue disease) in the USA, in Australia and in New Zealand limited live animal imports. Simmental breeding animals became expensive as the high demand could be met by neither imports nor the small national population (Daly, 1981; Sonn, 1985). Simmental imports to these countries had to go a long way round, such as via Canada or the UK, to reach their final destinations (Felius, 1995). BSE and Foot-and-Mouth disease in Europe in the late 1990s has influenced the exchange of breeding animals, semen and embryos for years as some countries refused to import from Europe. Similarly diseases and import regulations hinder the direct export of Simmental genetic material, for example from Southern Africa to Germany (Grupp, 2003). In the cattle sector, government policies are the main influence on breeding and gene flow. Regulations in cattle trade and breeding include market-regulating subsidies. Other government plans or projects influence gene flow by importing selected breeds. For example, in the early 1980s, the main imports of Simmental to Italy came from Austria and Switzerland. These imports were subsidised by the exporting countries and distorted the market, making it harder for other countries to export their breeding animals to Italy (Sonn, 1985). In Morocco a national plan to increase dairy production led to 3,865 Holstein heifers being imported from Europe and North America in 1986-87 to private dairy farms (Johnson et al., 1989). A government-funded project in Nigeria led to the introduction of 5000 N’Dama cattle from The Gambia between 1980 and 1983. They were meant as breeding stock for multiplication and dissemination in the country (Jabbar and Diedhiou, 2003). An initiative of Honduras’ President in 1973 led to the introduction of mainly Brahman breeding animals to be used by private ranchers and government AI centres (Blench and MacDonald, 2000). One of the most important “planned economy” schemes was the creation of the SMR (Holstein, Jersey, German Frisian) synthetic breed in the former GDR with 2 million cows (Meyn, 2005). In India, as another prominent example, the government started and supported programmes to create combination crossings in cattle relying on a considerable influx of exotics.

5.1.2.4 Pigs

As breeding pigs are mainly disseminated by a few international breeding and trade companies, particularly for hybrid pigs, in principle transport of any magnitude and distance is possible. However, practically, gene flow faces restrictions from veterinary regulations and welfare considerations. Although biotechnology can overcome some of these restrictions, generally breeding animals are transferred while AI, as deep-frozen ANALYSIS OF DRIVING FACTORS AND IMPACTS 71 semen, is used to maintain purebred lines in highly technical nucleus flocks. Animal welfare regulations influence transport costs drastically which has led to the installation of breeding units with multiplication systems in the destination countries. For example pig imports into Africa is considered irrelevant to gene flow because its strict health regulations force the big international pig breeding companies to use existing breeds. For a further example, Australia does not export because their strict health regulations prevented them from importing and so benefiting from the early European or North American advances.

5.1.3 Foreign aid and development projects

Foreign aid and development projects have focussed on introducing foreign breeds from northern to southern countries, or promoting crossbreeding efforts through appropriate funding for animal production. Many of those projects appear to have had no long-term impact, and so project strategies have changed. In the following chapter, the impact of development projects and the flow of funds are described from information available in scientific publications and project reports. The chapter is limited to examples, and there is no full regional coverage. The authors rarely had access to project reports, and where reports were available failures were often excluded. Therefore a balanced picture cannot be drawn. As a general trend, major movements of exotic breeds nowadays occur rarely in conventional development projects, but increasingly for emergency aid after natural or man made disasters.

5.1.3.1 Sheep

The need for economic development of sheep production in developing countries has created a demand for improved breeds to quickly increase productivity. For a long time temperate breeds were introduced into developing countries for pure- and crossbreeding purposes by national and international development and aid programmes (Gatenby, 1986). Due to complex reasons, such as genotype-environment interactions and socio-economic prevalence, these efforts were often unsuccessful. Nevertheless, foreign genes were introduced into local sheep populations, although it is not possible to verify the extent. Other development programmes for tropical regions considered the strong relation between sheep and its production environment. Examples are programmes based on importing improved tropical hair sheep breeds into Southeast Asia for crossbreeding with local breeds along with performance testing of all breeds under identical environments. For a well documented example, Cameroon hair sheep were introduced to Malaysia from Germany to be crossbred with local Longtail sheep, by a Research and Technology Development Project of the EU (Schäfer, 1998). For another, the Small Ruminant- Collaborative Research Support Program (SR-CRSP) introduced Barbados Blackbelly and St. Croix hair sheep into Indonesia to crossbreed with Sumatran wool breeds (chapter 2.3.3) (Gatenby et al., 1997a; 1997b; Gatenby et al., 1994; Gatenby, 1986). Developing the Dorper breed and various related synthetic breeds is an early example of a development program aiming to combine local hair sheep (Blackhead Persian) with 72 ANALYSIS OF DRIVING FACTORS AND IMPACTS imported sheep for the arid production environment. Here again the main driving factor was the commercial interest in the breed. During the history of sheep breeding it can be seen that sheep populations are often prone to a dramatic decline or complete loss during and after disastrous events such as wars. In Afghanistan for example, the effect of the war and two subsequent drought years in 1999 and 2000 led to a complete loss of almost all herds in the nomadic households in western and southern regions. Since these nomadic people depend on their stock this meant the complete loss of production resources. In events like this, foreign stock is imported to quickly rebuild former population sizes. The recent emergency aid to Afghanistan tended to import high-yielding breeds to replace the lost local stock (FAO-aktuell, 2000). However, the sustainability of these recent projects cannot yet be forecasted. Since 1973, indigenous animal genetic resources have been given more attention, subjecting traditional sheep breeds to various conservation and management programmes. These programmes affect gene flow in the sense that paying more attention to indigenous sheep breeds can stabilise or even increase their proportion in the local sheep population. Support programmes for indigenous sheep aim to establish resource flocks or herds. Here, spreading the genes of the indigenous breed is given preference (Ponzoni, 1992a). However, private farmers may resist participating in this kind of program if they have access to superior genetic resources. In the same context, some projects aim to improve indigenous genetic resources which may reduce the demand to introduce foreign breeds. Mueller et al. (2002) give an example of a large-scale genetic improvement project for the heterogeneous Corriedale population, which was installed in the Peruvian highlands to improve living standards of the Andean peasants by improving wool production. Corridale influence can nowadays be traced all over the Andean countries. However, the attempt to market the wool internationally failed completely, and farmers look for ways back to their older breeds with lower maintenance, higher milk yield and fertility. Nimbkar et al. (2002) reported a breeding programme to improve the Deccani sheep in India by crossbreeding it with two indigenous breeds. Arora et al. (2002) postulated that appropriate breeding strategies could be major tools to enhance mutton production and other sheep productivity. Indigenous breeds like Malpura, Muzzafarnagari, Madras red and Mandya sheep improved considerably through . Kosgey, et al. (2002) was one of a series of authors reporting that nucleus breeding had been proposed as a good strategy to improve livestock in developing countries. However, most such proposed programs have never been executed because of lack of long-term project funding. We can conclude that foreign aid and development projects can potentially start or stop gene flow depending on whether the aim is to replace or crossbreed indigenous stock with imported breeds, or to genetically improve local flocks. Where the imported stock is suitable for the production systems, development programmes tend to import tropical and subtropical rather than temperate breeds. However, the Awassi case study showed that commercially driven activities had far more impact than development projects. The spread of Merino sheep to developing countries as part of projects in the 1960s and 1970s did not bring sustainable economic benefits but did leave genetic traces in local breeds in many parts of the world. ANALYSIS OF DRIVING FACTORS AND IMPACTS 73

5.1.3.2 Goats

Over the past decades an increasing number of programmes started by government, local and international non-governmental organisations (NGOs) and international donor agencies have promoted or incorporated goat production in smallholder households to reduce poverty and develop economies. A few examples from publications and project reports illustrate the gene transfers brought about by such activities. Since information from development agencies is rarely accessible, the study cannot achieve global coverage or representation, but the following are considered typical cases. In Asia, Heifer Project International (HPI) has based many of its development projects on promoting goat husbandry. Pelant et al. (1999) list the goat types introduced to Asia between 1985 and 1997 but do not give volumes or breed details (Table 2). The UK Food and Agricultural Research Management Organisation (FARM-Africa) and the US Small Ruminant Collaborative Research Support Programme (SR-CRSP) have also followed this strategy. In Africa and Asia, the spread of dairy and meat goats was promoted by HPI and FARM- Africa with the “pass-on the gift” approach: any farmer or group receiving assistance in the form of animals, material, and/or training, will share or pass on at least an equivalent of the gift value to a neighbouring family or group. If animals were received, then the first female offspring had to be passed-on to the neighbouring family or group (Pelant, 2001).

Table 2: HPI projects with small ruminants in Asia (1985-1997)

Country Year initiated Types of small ruminants 1985 Meat goats, dairy goats People’s Republic of China 1989 Sheep Indonesia 1985 Meat goats India 1986 Meat goats, dairy goats, sheep Philippines 1986 Meat goats Thailand 1988 Meat goats Vietnam 1991 Meat goats Nepal 1994 Meat goats Bangladesh 1997 Meat goats Democratic Republic of Korea 1997 Dairy goats Sri Lanka 1997 Meat goats, dairy goats

Source: Pelant et al. (1999)

Since 1982, the Boer breed has been imported to Sri Lanka to upgrade the existing local population in the dry zone. This joint development programme with Germany claims to have had a significant impact in the five districts selected for the programme (Porter, 1996). While crossbreeds were heavier than local breeds under intensive conditions with supplementary feeding (Jeyaruban at al., 1997), the benefits of Boer goats and their crosses were not shown in smallholder extensive systems (Rajapaksha et al., 2001). 74 ANALYSIS OF DRIVING FACTORS AND IMPACTS

Experience has shown that development projects that are supported by governmental policies had a greater impact on the spread of specific breeds. This was the case in the Province in China, where in 1986 HPI introduced the Ya’an Dairy Goat project in collaboration with the Ya’an City Bureau of Animal Husbandry and the Ya’an Breed Improvement Station. Between 1986 and 1991, HPI provided 78 head of high quality Saanen does and bucks. 30 went to the station maintaining a high quality nucleus, and 48 went directly to 48 families. During this period, the total number of dairy goats rose from 6,750 to over 21,000 head. Through the “passing-on” approach 784 goats were shared with 716 additional families by the end of 1997. The main increase was due to a private and government-supported influx of additional dairy goats into the region, and the obvious success of the programme. By 2001, more than 20 provinces of China and two foreign countries had imported animals from Sichuan Province (Pelant, 2001). In the Indian Himalayan region, local NGOs actively promoted crossbreeding goats for milk and meat production (Chander et al., 2000). Table 3 shows more results from HPI projects in India. However, no breed information was given nor was the project’s impact assessed by external bodies. Self-assessments of projects can rarely be disconnected from self-interests related to image and marketing. Note also that this is only one of many small ruminant development aid projects operating in that region over decades.

Table 3: Number of goats distributed by HPI in India (1995-1997)

1994 1995 1996 1997 No. of goats including bucks distributed 2,622 1,813 1,524 5,959 No. of families helped 811 1,030 635 2,476 No. of families receiving pass-on animals 643 1,030 635 2,308 No. of goats passed-on to date 1,938 1,820 1,475 5,223

Source: Pelant et al. (1999)

In Samoa, in 1982 an IFAD livestock development project aimed to increase smallholder goat production. For that purpose, 372 goats, including goats from New Zealand and commercially reared goats from Fiji, were imported and distributed. By 1987, 60 goat units had been established. However, these units were not successful: there were insufficient market opportunities for goat meat, and a social stigma was attached to farmers who reared goats. Therefore, in 1989 the goat component of the project was abandoned (IFAD, 2004b). In 1971 Boer goats were imported into Kenya and were crossbred with the Small East African goat to improve meat production. In 1993 the Faculty of Veterinary Medicine, Makerere University of Uganda, imported 35 female and 5 male Boer goats from the Northern Cape Province of South Africa with the same objective (Nsubuga, 1996). However, as for many similar projects, the impact was not measurable. In 1999 four HPI goat groups were created in northern Tanzania. Toggenburg or Saanen does were imported from , Kenya or the USA and given to the group members; each group also received a buck, with specific mating rules. In 1999, of the 46 original goats reportedly 68 purebred offspring were passed on to other members. The group members ANALYSIS OF DRIVING FACTORS AND IMPACTS 75 and even non-member farmers were able to benefit from the buck to upgrade their local goats (de Haan et al., 2000). Similarly, FARM-Africa began its Dairy Goat Development Programme in 1988 in Ethiopia. Groups of 20 to 25 members were formed and received two local does on credit to be repaid in kind. In order to increase the production of local goats, each group received an imported buck of purebred dairy breeds - British Toggenburg and Anglo Nubian. Due to the success of the project, FARM-Africa has expanded this work to Kenya, Tanzania and Uganda after a positive evaluation of project success (de Haan et al., 2000). However, national and international research and development institutions in Ethiopia came to the opposite conclusion, and research and promotion was redirected to local goat populations.

5.1.3.3 Cattle

Cattle, particularly dairy cattle, have always attracted more interest than other species, and foreign aid and development projects are numerous in this sector. Well known examples are the International Fund For Agricultural Development (IFAD) programmes, the Heifer International programmes, the Balkan Emergency Farm Reconstruction Projects, FAO’s International Scheme for Coordination of Dairy Development (ISCCD), or FAO’s International Semen Donation Scheme. Often, to increase the national output from animal production in developing countries, improved temperate breeds are introduced and crossed with tropical dairy breeds. Many developing countries have sought loans from varies sources. However, due to complex reasons, such as genotype-environment interactions and socio-economic restrictions, efforts varied in their success. Projects can be successful where the introduced breeds are suitable to the climate and economies. Taking the humid tropics as an example for the most difficult climates, we can look at a project that introduced high-yielding dairy breeds into the tropical highlands of Kenya. Although that and the subsequent AI programme, supported by many donors, increased sustainable national dairy production, funds went missing and the programme collapsed finally in the 1980s (Philipsson, 2002; Meyn, 2005). An example of unsuccessful gene transfer into the humid tropics is the 2,400 dairy cattle introduced in the 1980s by the Philippine Government as part of an International Fund For Agricultural Development (IFAD) Project, meant for backyard production units. Although changes in government policy and unfavourable loan terms are given as explanations for the project’s failure (Arganosa et al., 1989; IFAD, 1988), the unsuitability of high-yielding dairy breeds for the humid tropics has to be regarded as the major reason (Meyn, 2005). Similar observations were made when Holstein cows were introduced to smallholdings in Sri Lanka, Vietnam and many others. These examples demonstrate the importance of politic stability and socio-economic factors for the success of projects. As another example, IFAD reports of a cattle restocking project as part of a livestock development programme in the Central African Republic in the 1990s. The objective was to raise livestock productivity and thereby alleviate herder poverty, but the restocking was abandoned due to political instabilities. Other difficulties concerning restocking programmes that are discussed by IFAD (2004a) are that they tend to be politicised and that it is difficult to reach the identified target groups. In Tanzania, a great number of heifer-in-trust-programmes by several organisations were implemented with 700 heifers procured. Repayment levels were so low that the programmes were not 76 ANALYSIS OF DRIVING FACTORS AND IMPACTS sustainable. It has been argued that the main problems are due to participants’ inexperience of the transferred pregnant heifers, which generally require more management compared to their traditional cattle (Afifi-Affat, 1998). However, it can also be argued that technical solutions should fit prevailing local conditions and not vice versa. Based on these problems of sustainable dairy cattle introduction, an approach was derived to procure dual-purpose breeds and produce crossbreds of exotic and local breeds. Heifer International has implemented projects to procure Simmental cattle in Ukraine (Heifer International, 2004b), China (Heifer International, 2004d), Poland (Heifer International, 2004a) and Kosovo (Heifer International, 2004c). The projects were meant to improve agricultural production and so contribute to the nutrition and income of the beneficiary families. Emergency Farm Reconstruction Projects on the Balkan after the conflicts in this region led to imports of Simmental breeding animals, mainly pregnant heifers (Cossee, 2003). Institutions like the World Bank, FAO, and IFAD were involved (World Bank, 2003). IFAD (1984) reports of a crossbreeding scheme in Guatemala with high shares of Brahman compared to Brown Swiss. The recently favoured funding improvements to local breeds began because of failures to introduce sustainable exotic breeds. However, access to information about failures and successes of development projects is difficult as donors protect their sources to avoid negative images and so lose funding. Nevertheless, projects have increased gene flow mostly north to south, and have contributed to the world wide spread of cattle breeds from North America and Europe, particularly the Holstein breed.

5.1.3.4 Pigs

In Vietnam exotic breeds have been intensively promoted and imported as part of government strategies. The impact of these state-run breeding stations on gene flow has been considerable, regardless of whether the genetic material stemmed from colonial times or was exchanged with other socialist countries. National development projects and breed distribution projects have distributed the genetic material. Additionally, the Vietnamese research and education systems supported the use of exotic breeds and crossbreeds, advanced management techniques, and promoted AI by building AI stations. Since the late 1980s, with the start of the open door policy, government and non- government organisations have helped to introduce and distribute higher-yielding pig genotypes in both small and big projects. Consequently most indigenous breed populations are declining and the majority of breeds are vulnerable or critical, some even facing extinction. Only a small number of breeding centres and research institutes keep local breeds for crossbreeding and conservation. As the breeding organisations are not centralised in Vietnam, genetic improvement, breed replacement and conservation are not uniform across regions (ASPS, 2002). In some Caribbean countries, local pig populations have become virtually extinct after massive introduction of North American breeds as part of development aid. However similar projects in the Andean countries in the mid 1990s showed only little penetration. A similar situation occurs in Africa with regard to the impact of European breeds. In Asia, however, development projects have much less impact than commercial activities, creating a totally different situation. ANALYSIS OF DRIVING FACTORS AND IMPACTS 77

5.1.4 Research and commercial interests in specific genes

The interest in specific genes affects gene flow as it prompts the flow of these genes from their countries of origin to all continents where the commercial or research interest exists.

5.1.4.1 Sheep

Research and commercial interest in specific genes in sheep can be summarised as prolificacy, disease resistance and adaptability to certain production characteristics. The aim is to identify these genes and to exploit their productivity potential. Major genes are of major interest to realise quick genetic improvements, making sheep pioneers in their research and commercial use. Improved prolificacy has gained increased global interest since the re-orientation towards meat production in sheep breeding. The Booroola gene as a major gene affecting prolificacy is the best example for identifying and disseminating a specific gene through crossbreeding, as is its global success in quickly improving prolificacy in several breeds world wide (for details refer to the sheep global study in Annex 9.2) but also for being rejected if not suitable under specific natural or socio-economic conditions. Fahmy (1996) compiled information on the spread of prolificacy genes in sheep breeds. As well as the Booroola gene, other prolificacy genes have been identified e.g. in the Chios breed in Greece, the D’man sheep of Morocco, the Barbados Blackbelly of the Americas, the Javanese thin-tailed and Javanese fat-tailed sheep of Indonesia, and prolific sheep of China. Breeds with recently discovered genes for prolificacy are the , and the Icelandic sheep. Less known in this respect are the Belle-Ile sheep of France, the and the Teeswater of England, the Flemish Landrace of Belgium, the Galician of Spain, the Garole of West Bengal, India, the Imeritian of the Republic of Georgia, the Olkuska of Poland, and the St. Croix on the Virgin Islands. Davis et al. (2002) confirmed the presence of the BMPIB receptor related to the Booroola gene in the Indian Garole and Javanese thintail breeds. The Javanese thintail breed has gained particular research interest because the single gene for prolificacy is segregated in the population and so can be used to develop strains differing in reproductive performance, to exploit for different intensities in different production systems (Ponzoni, 1992b). The D’man sheep from Morocco is a minor breed in the country, but it has attracted attention because of its exceptionally large litter size, early puberty, and short lambing interval. However, it appears that in the D’man the high litter size is transmitted additively (not via a single gene of large effect). Since the breed evolved in a very hot environment, apart from being valuable itself, it could be useful for crossing with other breeds in similar environments (Ponzoni, 1992b). The global research and commercial interest in improving disease resistance and adaptability brings tropical and subtropical sheep breeds to the centre of interest. Adapting sheep to the tropics requires losing some fleece and improving parasite and disease resistance to survive and reproduce under these conditions. Segregating a single major gene responsible for expressing hair coats is an example of identifying a specific single gene that research and commerce can use to reduce woolliness in populations of tropical and subtropical countries. Additionally, reducing woolliness has gained interest in temperate countries when the costs of shearing exceed the return from wool. Many tropical 78 ANALYSIS OF DRIVING FACTORS AND IMPACTS hair sheep breeds also have genetic potential for a high lambing rate, and the potential for more frequent lambing in tropical regions is apparent. (Shelton and Figueiredo, 1990). Researchers are interested in the Djallonke hair sheep of West Africa because they are able to survive and reproduce in trypanosomiasis affected areas (Goossens et al., 1999; Goossens et al., 1997; Ponzoni, 1992b; Mavena, 1987). Fitzhugh and Bradford’s (1983) argue that most American hair sheep belong to the same “genetic type”, so this interest extends to American hair sheep as well. Great research interest has developed in countries with intensive production systems in the field of resistance to helminth infections with Haemonchus contortus. This is not inherited via a single gene, due to the increasing resistance of breeds to anthelmintics. Of particular interest is the resistance of Djallonke (Goossens et al., 1999), Barbados Blackbelly (Yatwinski et al., 1980), Red Maasai (Mugambi et al., 1997; Preston and Allonby, 1979), and St. Croix hair sheep (Gamble and Zajac, 1992). Sansthan and Köhler-Rohlefson(2005) add the Javanese Thintail, the Garole, and the Florida Native to the list. The authors also find the Javanese Thintail, Indonesian Thintail and St. Croix resistant or tolerant to Liver Fluke, the Red Maassai resistant to Rift Valley Fever and Maedi visna, and the Wensleydale resistant to Scrapie. In most of the cases reported, the genetic foundation of the traits is still unclear. Research interest has arisen in the Prion Protein Gene (PrP) for resistance to or susceptibility for Transmissible Spongiform Encephalopathies (TSE). ICAR reports PrP genotyping in 7 countries to look for gene frequency changes after selection for preferred allele types. Due to the quick genetic improvements that are now feasible and their easy introduction into new populations, identifying major genes responsible for superior individuals or populations is given preference in sheep research and commercial interests. Improving productivity of many breeds world wide, by introducing prolific genes or reducing woolliness for example, is evidence of gene flow in the past. Breeds from countries with traditional sheep production systems move into the centre of interest as their genetic diversity is greater than in countries with intensive sheep farming. It has been shown that interest in identified breeds goes far beyond their national importance. Exploitation exists in the country of origin, in countries with similar production environments and globally.

5.1.4.2 Goats

There are scientific rather than commercial interests in targeting major or single genes in goats that determine milk composition as in cattle, prolificacy as in sheep, and disease resistance as in most domestic animal species. With regard to disease resistance, the West African dwarf is tolerant to trypanosomiasis and gastro-intestinal nematodes, and the Small East African to gastro-intestinal nematodes. The research interest focuses on the commercial use of such genes.

5.1.4.3 Cattle

Marbling is a main determinant of beef prices in international trade. In Australia gene markers have been identified to predict marbling potential. They have been commercially marketed and patent protection has been sought. There are also efforts to identify gene ANALYSIS OF DRIVING FACTORS AND IMPACTS 79 markers for parasite resistance and net feed efficiency (Allen, 2002; CSIROnline, 2004). Interests to identify and exploit major genes determining milk composition in dairy cattle exist. Unlike the situation in highly prolific species of poultry or sheep, major genes in cattle cannot be exploited by conventional breeding methods of recurrent , but depends on the feasibility of gene transfer and the industrial process of legalising genetically modified animals, leaving very few companies able to compete. Disease resistance in local breeds is reported for the East African Short Horn Zebu for Brucellosis, the Small East African Zebu for East Coast fever, the Brahman for Haemoparasites and ticks, the Orma Boran for trypanosomiasis, and the N’Dama for Cowdriosis, Dermatophilosis, gastro-intestinal parasites, ticks and trypanosomiasis (Sansthan and Köhler-Rohlefson, 2005). Research interest in disease resistance is high, however, commercial exploitation in the near future is even more doubtful than for the examples mentioned above. So far, many claims for disease resistance of local populations have not been scientifically proven or specified.

5.1.4.4 Pigs

Consumer preferences play an important role in the demand for specific genes to serve them. For example, some Taihu breeds are imported for specific fertility genes by France, the Netherlands, the UK and USA, but no commercial use has been seen so far. Some interest in Vietnam’s “pot-bellied” pigs as pets has been reported, and they have been used to create experimental mini pig populations in northern countries. There has also been recent scientific interest in Coli-receptor genes in pigs. Some indigenous livestock breeds with genetic disease resistance are summarised by Sansthan and Köhler-Rohlefson(2005). They describe local indigenous pig breeds from Democratic Republic of Congo, Mozambique, Angola, and Sudan that display moderate to high resistance to African swine fever. Also, the I-pig from Vietnam that displays some (but unmeasured) resistance to Foot-and-Mouth disease. None of these examples have been of commercial interest. Simlarly, 9 purebreeding pigs were given by China as a gift to France, and genetic material from trials with these pigs were spread to other European countries and the USA, mainly for research purposes. As already mentioned in chapter 5.1.1, unique germplasm from Asian breeds has played a vital role in commercial pig breeding programmes of European and American companies.

5.1.5 Suitability of breeds for prevailing production systems

The suitability of introduced improved genotypes for prevailing production systems determines the impact of the imported breeds on the existing breeds. If the new genotype is very suitable, they often replace the local stock. If it proves unsuitable, it will eventually disappear with virtually no impact. Depending on the production system, optimal levels of genetic contributions from the respective breeds and optimal breeding schemes need to be considered before recommending gene flow plans (Cunningham and Syrstad, 1987; Cunningham, 1991; Valle Zárate, 1996; Blake, 2004). Recommendations to promote or discourage gene flow have been based upon genotype comparisons in place. Appropriate indicators are a key feature to determine the suitability of genotypes for a given production system with its natural, economic, social and cultural conditions. Usually biological or economic 80 ANALYSIS OF DRIVING FACTORS AND IMPACTS efficiencies have been used. However, breed comparisons should also take farmers preferences for certain breeds or traits into account. Determining appropriate indicators in that case however is less developed than for biological or economic indicators, and frameworks to analyse developing world systems are not nearly as well in place as developed world systems (Gibson and Pullin, 2005). Indigenous breeds usually fulfil a wider range of functions for their owners. The economic significance of non-market functions and even non-use is being researched (Drucker et al., 2005).

5.1.5.1 Sheep

Examples of introduced foreign breeds proving unsuitable for their new production environments are numerous in the history of sheep breeding. For many decades, many temperate breeds introduced into the tropics or subtropics have been unsuccessful due to genotype-environment interactions (Horst and Reh, 1999). The genetic impact of these imports on existing breeds was therefore small. Iniguez’s review (1998) of community breeding programmes for small ruminants in the Andean region showed that development programmes in the region tended to use exotic germplasm indiscriminately without appropriate breed comparisons. Another example is the widespread cross of the non-seasonal Criollo and the seasonal Corriedale. Improved germplasm was introduced based on its merit in its place of origin, and was not properly evaluated in breed comparisons in the region. There was therefore little information on the suitability of the breeds for the prevailing natural and socio- economic conditions. In Peru, some government agencies and NGOs are still distributing purebred rams, particularly Corriedale, to upgrade Criollo/Mestizo sheep. Because community producers were not adequately involved, most programmes only lasted while funding was available or government commitment remained. Similar experiences with the transfer of breeding stock to developing countries are reported in the Awassi case study. However, there are also numerous examples where introducing improved foreign breeds has significantly improved productivity of sheep farming when the introduced breeds are suitable for the new production environment. Establishing foreign genes in the local populations is regarded as proof of their suitability. The whole history of the Merino breed is a vivid example for the global impact of an improved breed on numerous sheep populations in the world, based on the breed’s suitability to a broad range of environments. Establishing foreign wool breeds and their crosses in the environment of the high lands of East Africa, South America and India are examples for the suitability of temperate breeds to environmental conditions of tropical or subtropical countries. However, they have not always proved their fit to the socio- economic framework. The Awassi case study shows that today’s Awassi populations are more numerous in countries where they have been exported to and fitted to the production system than in the country of its origin. Shrestha (2005) reviews the creation of 443 reported composite breeds or lines in 68 developed as well as developing countries of the world during the last centuries, based on introduced foreign breeds. It proves the significant impact of foreign gene flow on breeding directed at improving productivity of sheep husbandry. Although the impact of the breeds was not given, many of them did not prove competitive and sustainable. However, some examples of successfully established composite breeds, identified from the ANALYSIS OF DRIVING FACTORS AND IMPACTS 81 list, support the conclusion. Developing and establishing the Dorper breed as the second largest breed in South Africa and its popularity in many other countries is the most striking example. The increase of the percentage of imported breeds of the total population from 4.1% to 15% from 1976 to 2000 in Poland gives another example of suitable and sustainable gene transfer, besides reflecting the considerable impact of political and economic framework conditions for the spread of breeds. The growing awareness of the importance of suitability of breeds in their new production environments has led to two consequences: one is to increasingly consider introducing improved breeds from similar environmental conditions and production systems, and the second is to performance test the different genotypes in the initial phase of gene transfer. The Improved Awassi for example showed, on station with intensive management under subtropical conditions of the destination countries, higher milk and mutton production compared to indigenous breeds. However, under the tropical conditions of India and Ethiopia, and the low-input production systems of Eastern European countries, the Improved Awassi could not maintain this performance. Schoeman (1996) gave a comparative assessment of Dorper sheep in different production environments and systems and concluded that in the extensive areas of South Africa, the Dorper generally had superior reproductive and growth performance to woollen and other indigenous breeds. It also compared favourably with specialised mutton breeds. However, the Dorper proved to be unsuitable for commercial use in the African tropics. Literature reviewed by Schäfer (1998) gives evidence of the suitability of hair sheep crosses to tropical conditions based on performance comparisons, where possible directly comparing with purebreds. An example of evaluating performance of hair sheep crosses, from crossbreeding Barbados Blackbelly and St. Croix hair sheep from the USA with Sumatran wool sheep breeds in Indonesia, is presented by Gatenby et al. (1997a and 1997b) and Gatenby et al. (1994). Several sheep crosses were evaluated in many areas of Algeria, with breeds imported from Europe, Australia, New Zealand and recently from the Middle East. An early crossing between French Merino rams and Ouled Djellal ewe gave a new strain, named Taadmit, and it now exists in the highlands. This breed has good conformation and good wool quality. Crossing experiments carried out in some state farms indicated reproduction and adaptation problems for the imported breeds. Growth rates were also very variable. It was concluded that introducing imported breeds is not recommended where breeding environments are not controlled (Benyoucef, 1992). Changing economic conditions may also affect the suitability of the prevailing breed to the new conditions, and subsequently may lead to re-orientating the breeding. For example the new market economy in the former GDR led to the re-orientation from wool to meat: main breeds were selected differently, local breeds disappeared and mutton breeds introduced. It also has to be taken into account that shepherds may continue to keep their local breeds despite the availability of suitable improved exotic breeds simply for tradition. This observation has been made across continents, mostly in traditional cultures with shepherds rather than cattle breeders who seem to be more open to new experiences. It is concluded that if the introduced breeds are suitable for prevailing environments and production systems, gene flow contributes significantly to improved productivity of sheep farming through the introduced improved breeds. However, establishing suitable foreign 82 ANALYSIS OF DRIVING FACTORS AND IMPACTS genotypes in local populations also threatens local breeds either through direct interference or through market competition.

5.1.5.2 Goats

The general adaptability of the goat enables this species to be reared in both highly intensive and very extensive production systems. In highly intensive production systems, there are the high-yielding, highly prolific and regularly breeding dairy types: the Alpine, Saanen, Toggenburg, Anglo Nubian and Damascus with milk yields of 800 to 1,000 litres per lactation. At the other end of the spectrum, one can find all over the world many lower performing breeds which excel by adapting to harsh environments, resisting diseases, and thriving under low-input systems. In the global gene flow of goats (Annex chapter 9.3), and particularly the Anglo Nubian case study (Annex chapter 9.7), examples are given of successes and failures of goat transfers, depending on their fit to environments and goat keeping systems. Dryland grassland is susceptible to invasion by bush. Because of their feed preference for bush, goats are widely used to control bush encroachment. Traditionally this is practised in Mediterranean countries, South Africa, Australia, New Zealand, Latin America and the southern United States. In the USA the breed used is called “Spanish goat”. They have been exported from Texas to other states to manage vegetation. In California goats are used near urban and forested areas to create firebreaks by out shrubs from adjacent land strips (Glimp, 1995). In the mid-west and north of the USA and Canada, Angora goats were used in the 19th century to prepare new farming land by feeding on bush. Many Mediterranean range goat production systems are located in the most difficult environments where goats graze the especially harsh areas while other ruminants are allotted to better pastures at the same location. This is the case of Angora goat flocks in Turkey (El Aich et al., 1995). Due to the great performance differences between temperate and tropical/subtropical goats, crossbreeding is an opportunity to use these differences to develop synthetic breeds. Examples from Kenya, Burundi, and Malaysia show it is necessary to define optimal gene share of the exotic breed related to production purpose and production systems (Horst and Reh, 1999).

5.1.5.3 Cattle

The spread of the Holstein breed is one of the most significant cattle gene flow events in history and a vivid example of the global impact of this high-yielding dairy breed where milk and beef prices favour dairy production, and where the climate is suitable. The latter is the case in the whole temperate zone. Purebred Holstein are also expanding into the dry tropics and subtropics and actually achieve their highest yields in Israel and California. In the humid tropics efforts to introduce Holstein cattle have been of limited success. In a global analysis of Holstein cattle imported to warm climate regions (30° latitude north and south) it was stated that Holstein cows had on average 125 kg less weight, showed 30-50% lower milk yields, displayed 10-15% longer calving intervals and had 0.5 to 1.2 lesser calves in their lives (McDowell, 1992) compared to Holstein cows in temperate climates. For a summary of some studies of the Holstein breed in different environments see Annex ANALYSIS OF DRIVING FACTORS AND IMPACTS 83

9.1 A 48. (McDowell, 1985; Blake, 2004; Brahmia and Ben Dhia, 1990; Chagunda, 2000; Ikiror, 2001; McDowell, 1992; Nassuns-Musoke, 2002; Syrstad, 1991). Comparing different exotic breeds and their crosses in the tropics, it was found that the Holstein was not necessarily superior (Rege, 1998). In a review in 1990 with the main focus on tropical Latin America it was stated that losses in imported European dairy breeds were generally so high that they were unable to produce sufficient replacements to maintain herd size. A fact borne out by the need for continuous imports to keep herd sizes of European origin constant (Vaccaro, 1983) cf. (Philipsson, 2002; Blench, 1999) makes similar statements for European dairy breeds introduced to Africa: Even under high management conditions these breeds could barely be sustained economically. Exceptions are known from Kenya and South Africa, where some high-input farms produced for urban markets (Blench, 1999) or Uganda where a Holstein population could successfully develop over decades in a region with a suitable climate (Nakimbugwe, 2005). Local cattle are crossed with Jersey or Holstein cattle in Zimbabwe for commercial milk production (McDowell et al., 1996). It has to be added that even in countries with no traditional milk production, strong national subsidy systems may lead to the development of a national dairy production, as for example in Vietnam. However, generally national subsidies are not maintained over the long run and production costs rise above world market levels and undermine sustainable dairy production as in Sri Lanka. Dairy production surviced in countries like India, Lebanon and Syria due to maintainance of national subsidies. When comparing the suitability of improved dairy breeds for new production systems, the lack of genotype-environment interaction analysis led to the foundation of the International Bull Evaluation Service (Interbull). This organisation helps to prevent temperate sires being genetically overestimated when used in tropical and subtropical countries. Although virtually all data come from intensive dairy systems that do not provide breed value assessments in most developing countries, it has given importers better information for their decisions (Smith and Banos, 1991; Banos and Sigurdsson, 1996; Zwald et al., 2003). There are many breeds suitable for diverse purposes in the different ecozones: B. taurus breeds from Britain and the European Continent are suitable for the temperate zone, B. taurus x B. indicus crosses and Zebu breeds, Sanga breeds and other indigenous breeds are suitable for subtropical areas, and specially adapted taurine breeds like N’Dama are suitable for tropical environments. Rege (1998) 80 studies of crosses of Bos taurus and Bos indicus cattle breeds in the tropics. Milk yields increased with up to 50% Bos taurus gene contribution, with a less pronounced trend for higher contributions. If milk yield was related to calving interval there was no clear trend at all. Age at first calving decreased with increasing shares from Bos taurus while calving interval was lowest for halfbreds and increased again substantially with higher taurine gene proportion. For a graphical depiction see Figure 8. Similarly Syrstad (1991) reviewed 54 studies and further differentiated for Bos taurus breeds and acknowledged breed differences, which were masked in the combined analyses. 84 ANALYSIS OF DRIVING FACTORS AND IMPACTS

Figure 8: Performance of Bos indicus, Bos taurus and their crosses in the tropics

3,000 470

460 2,500

450 2,000 milk yield (kg) 440 age at first calving, days calving interval, days 1,500 milk yield, kg 430 calving interval, days age at age first calving, days 1,000 420

500 410 0 1/8 1/4 3/81/2 5/8 3/4 7/8 1 (F1) proportion of Bos taurus contribution

Source: data from Rege (1998)

In the tropical and subtropical zone of Asia emphasis is on crossbreeding for milk production. As many efforts to establish purebred temperate dairy breeds have failed, India, Pakistan and Bangladesh have started crossbreeding taurindicus cattle for milk. In India, crossbreeding with exotic cattle breeds began in the late 1800s and it became an official policy at the turn to the 20th century. In the 1920s and 1930s European dairy cattle breeds (Holstein, Brown Swiss, Jersey, Red Danish, Ayrshire, , Guernsey) were introduced and used on Red Sindhi, Sahiwal, Tharparker, Gir and Harina in military farms. Continued crossings of mainly Holstein with Sahiwal revealed that up to an exotic share of about 63% performance levels increased - given that good management was secured. Based on these results between 1961 and 1975 exotic breeds were imported to produce semen which was used in countrywide crossbreeding programmes. Further crossbreeding schemes are reported from the 1980s and 1990s with Sahiwal and Holstein to form a new synthetic breed - the Frieswal - which was meant to be spread over the whole of India. Similar schemes are documented using other breeds, for example Brown Swiss, for crossbreeding in India. In the 1990s India was the largest user of crossbred cattle in Asia (Felius, 1995; Katpatal, 1977; Mudgal and Arora, 1994; McDowell et al., 1996). Mathias and Mundy (2005) summarise the percentage of crossbreeds with indigenous cattle in Asian countries, and cite that Thailand with 84% crossbreeds of the total cattle population has the highest percentage, followed by Sri Lanka (29%), Pakistan (7.5%), India (7%), and Bangladesh (2-3%). Crossbreeding local cattle with imported breeds (Holstein, Jersey, Sahiwal) and among these breeds (Syed et al., 1996) and of Sahiwal with Swedish Red and White breeds is also reported from Pakistan (Ishaq et al., 1981). Malaysia has imported large numbers of Sahiwal x Holstein crosses from Australia and New Zealand (McDowell et al., 1996). In some cases they were considered to be financially viable in smallholder production (Wan ANALYSIS OF DRIVING FACTORS AND IMPACTS 85

Hassan et al., 1989); in other cases widespread lactation failures were reported (Murugaiyah et al., 2001). Madalena et al. (1990) included management intensities in their comparisons of different crossbreds in Brazil. They found milk yields increased in above 50%-taurine-crosses in intensive management while in low management conditions the trend was unclear. This again underpins the importance of gene × environment interactions. The Simmental breed, originally a triple-purpose breed, has proved suitable for various production systems and climatic conditions. The use varies from country to country due to milk prices and the milk-beef-price ratio. While in Central and Eastern Europe it is used as a genuine dual-purpose breed, in the New World, Northwest Europe, Africa and Latin America it is mainly used for beef. Simmental genetic material has been used in crossbreeding schemes either for industrial crosses, to develop new composite breeds or upgrade local cattle breeds. Simbra became the most widely distributed synthetic breed to which Simmental has made substantial contributions. In Southern Africa (South Africa, Namibia, Botswana) the number of Simbra animals and Simbra breeders grew constantly from the mid 1990s until 2003 while the numbers in the USA and Australia were roughly constant (World Simmental Fleckvieh Federation, 2002). In Brazil the synthetic breed of Simmental and zebu breeds is the Simbrazil. The Brahman, adapted to hot climates due to its superior thermoregulation, and the Santa Gertrudis (3/8 Brahman and 5/8 Shorthorn) breeds had the most influence on beef cattle breeding in the tropical and subtropical zones of the New World (Brahman Cattle Breeders Society South Africa, 2004; Hernandez-Ceron et al., 2004). But naturally, they were not successful as subsistence milk suppliers for pastoral herdsmen. In the Philippines for example, genetic improvement through a nucleus breeding programme of imported purebred Brahman and Simbra cattle successfully reached rural village level by crossbreeding with local cattle by AI or natural mating (Boadle et al., 1997; Loculan, 2002). In general, breeding in marginal areas which had no tradition of beef production was made possible by introducing tropical cattle breeds such as the Brahman and its derivates In Latin-America, at the end of the 19th century when the British beef market was made available by the new refrigerated ships, Criollo cattle were upgraded to British beef breeds (Shorthorn, Hereford, Aberdeen Angus) or to zebu breeds from India. New crosses of imported cattle breeds have been continuously tried out and changed with market demands, varying management intensities and agro-ecological-regions (Bondoc et al., 1997; Madalena, 1977; Mandibaya et al., 1999). When cattle producers need multi-purpose cattle, or product prices do not justify costly inputs, other breeds and breeding methods have to be used. Here, the Sahiwal and its crosses with temperate zone dairy and dual-purpose breeds are promising. In Kenya there has been a great number of crossing trials of Sahiwal with the local East African Zebu and Bos taurus breeds (Table 4). From Kenya Sahiwal genetic material has been exported to the whole continent of Africa. 86 ANALYSIS OF DRIVING FACTORS AND IMPACTS

Table 4: Sahiwal crosses with other cattle breeds in Kenya

Breeds crossed with Sahiwal in Kenya Source Jersey, Frisian Meyn and Wilkins (1974) Ayrshire Kimenye and Russell (1975) Ayrshire Trail and Gregory (1981a) Ayrshire, Brown Swiss, Simmental, Friesian Trail and Gregory (1981b) East African Zebu Cunningham and Syrstad (1987) Ayrshire, Brown Swiss Thorpe et al. (1994) Ayrshire, Brown Swiss, Friesian Kahi et al. (1999) Ayrshire, Brown Swiss, Friesian Kahi et al. (2000) Friesian Mpofu and Rege (2002) The breed was formed from 1860 onwards on the Virgin Islands from imported N’Dama and (Felius, 1995). Also in the 19th century, in India and Sri Lanka, Taylor and Hatton cattle were developed from crossbreeding trials (Horst and Reh, 1999). In Western Africa there were crossbreeding programmes with N’Dama in the second half of the 20th century: in Côte d’Ivoire with Jersey imported from Great Britain (Letenneur, 1978) and Simmental from Germany (Meyn, 2005), in Sierra Leone with Sahiwal imported from Kenya and in Togo with German Yellow to form a new synthetic - the Avétonou (Felius, 1995). The Foula in Guinea Bissau, the Djakoré and Bambey in Senegal and the Bambara in Mali are crossing products of N’Dama and Fulani zebu (Felius, 1995). In Ghana Ndagu was bred from N’Dama and Sokoto Gudali imported from Nigeria (Felius, 1995). The Baoulé in northern Ivory Coast, in southern Mali and southwest Burkina Faso is also supposed to have received some gene flow from N’Dama. The Baoulé was exported to Liberia, Gabon, Togo, Central African Republic and Zaire (Felius, 1995). All these are examples of attempts to develop composite cattle breeds with high production and adapted to specific conditions. Only in a few cases have the composite breeds been further transferred and introduced to similar production conditions. The Australian Milking Zebu was developed in Australia in the 1950s by crossing the Zebu breeds Sahiwal and Red Sindhi with Jersey. Gene proportions ranged from 3/8 to 4/8 Zebu and 4/8 to 5/8 Bos taurus. The Australian Milking Zebu has also spread to Southeast Asian countries like Malaysia and Bangladesh. In Java, the Grati cattle was developed by crossing Dutch breeds with local Madura and Banteng cattle. In Brazil, the Pitangueira cattle is a cross of the imported Guzerat cattle from India and the Red Poll cattle with a 5/8 Bos taurus gene proportion. Horst and Reh (1999) also summarise beef cattle breeds developed through crossbreeding schemes. Beside the already mentioned Santa Gertrudis and cattle in the Southern USA, the (5/8 Angus and 3/8 Brahman), and the Charbray (Charolais and Brahman) also developed there. In South Africa, the Droughtmaster (1/2 Shorthorn and 1/2 Brahman), the (Hereford and Brahman) and Belmont Red (Hereford, Shorthorn and Afrikander) is mentioned. An example of an attempt to improve a breed and exporting it from developing to other developing countries based on similar ecosystems and production conditions is the nucleus breeding programme established for Criollo cattle. It led to initially successful introducing ANALYSIS OF DRIVING FACTORS AND IMPACTS 87 improved Criollo cattle from Nicaragua to Mexico and from Costa Rica to several Latin American countries. The most widespread distribution was reported in areas where milk producers were already organised and established institutions could be used (Alba, 1978; Felius, 1995)

5.1.5.4 Pigs

Commercial pig production differs from ruminant in that the environment is frequently shaped to the needs of the pigs. Adaptability problems are therefore in many cases reduced. Additionally commercially owned breeding enterprises often have regional structures in which breeding pigs are adapted to the environment. Therefore, exotic genotypes and their crossbreds are often better than indigenous breeds. Subsequently, commercial pig stock even in tropical and subtropical climates, for example South America, mostly consist of commercial hybrids and exotic purebreeding lines used as maternal stock to cross with terminal sires to produce pork. However, it was concluded that genotypes from temperate regions and their crosses face severe problems with traditional feed and management in developing countries, resulting in poor reproductive performance and high mortality. Furthermore to avoid loss of heterosis in consecutive generations, continual inflow of exotic germplasm is necessary which makes the success of small-scale crossbreeding schemes doubtful. Institutional, organisational and technical support are required for these crossbreeding schemes to succeed (Agbagha et al., 2001). In most cases breeding with exotic breeds in smallholder conditions was neglected after the projects completed. Without further access to exotic breeds, the impact of selective crossbreeding on the local populations vanishes from generation to generation but already had diluted local breeds. The numerous examples of successful breed transfers and breed combinations that give local farmers, by increasing performance potential, opportunities to develop their production systems and livelihoods will never be adequately summarised.

5.2 Impact of gene flow on economic development and poverty reduction From the main findings discussed in chapter 5.1, the following conclusions are made about the impact of gene flow on economic development and poverty reduction. Developing countries display a demand for new breeds of all four species to increase productivity of farming. With biotechnology and better transport, genetic resources have become readily available and accessible across the globe. However, the quantity of genetic material transferred is not important to the success of gene flow. Instead it is its suitability to the new environments, economies, cultures and societies as discussed in chapter 5.1.5. If the introduced improved breeds are suitable for the prevailing conditions and production systems, gene flow contributes significantly to the country’s economy and reduces poverty by improving on-farm productivity. This is true for dairy cattle but may also include increased red meat production as reported for pigs, sheep, goats and beef cattle. The importance of gene flow for economic development is particularly evident where the local breeds have a low production potential and so cannot be quickly improved through within- breed selection. When evaluating such programmes, the costs to implement selection schemes and the relatively slow progress are often underestimated. On the other hand, as 88 ANALYSIS OF DRIVING FACTORS AND IMPACTS imported gene flow often involves specialised and concentrated production the poorer rural population may not benefit from it. The greater the success of the introduced breed, the greater the threat of losing biodiversity in the national population. Many marginal areas, such as deserts, scrublands and mountainous areas, can only be exploited by locally adapted breeds. However, nobody has ever attempted to replace local breeds by exotics under these extreme conditions. If the new breeds add to the diversity of national populations, gene flow offers the chance to quickly adapt to new production options. Particularly, in rapidly developing parts of the world, gene flow vitally enhances production to meet the population’s growing demand for red meat. Milk production and marketing schemes may directly benefit smallholders. Production can be further improved if this development goes hand in hand with a change from extensive to intensive production systems. These intensive production units will compete through the market with extensive production systems, so threatening the livelihoods of smallholders. Negative effects of market competition between big units and smallholders can arise if governments help to establish large-scale livestock production through laws and incentives that exclude smallholders (Mathias and Mundy, 2005). However, it is a political decision how national resources are allocated and whether or not efforts are made to ensure poor farmers participate in the development process on or off their farms. Left to survival strategies, they are better off with their local breeds than without them. To get out of poverty, relying on local breeds will hardly ever suffice. Restricting access to advanced technology, here specifically access to high performing exotic breeds, is likely to hinder self-sustained development. However, protecting the developing local breeding organisations and avoiding hidden subsidies to exotic competitors may help local livestock breeders to participate in economic development. Measures to counteract monopolisation in the breeding organisations and access to genetic resources may also help. The choice of technology, here the access to breeds and genetic material, cannot replace political decisions. In conclusion: if a programme cannot be implemented to keep gene proportion of the exotic breed at appropriate level, introducing exotics is dangerous. Gene flow may also contribute to niche market production, permitting or even encouraging the use of foreign breeds with characters specific to local market preferences. Niche markets provide a promising chance to survive, particularly in sheep farming which otherwise faces severe reductions or even lost profitability. On the other hand, native genetic resources may contribute as much to niche market production or other production purposes which also conserves national genetic resources. Fat-tail sheep e.g. cover a niche market in Central Asia, where meat is produced for the Arabian market. A particular importance of gene flow is given when livestock populations are reduced, for example after wars. When effective breeding work is limited and the basis to restore and develop future populations is lost, gene flow is the only chance to rebuild the stock from imported genetic material. However, a decision has to be made whether the rebuilding should introduce high performance breeds from temperate climates or local breeds from neighbouring regions. If the imported breeds are unsuitable for prevailing production systems, the impact of this gene flow on population composition is theoretically self-limiting. However, the economic losses for smallholders can be severe, if development projects or governments continue to ANALYSIS OF DRIVING FACTORS AND IMPACTS 89 promote or subsidise these genotypes. The losses are caused by diluting or replacing the indigenous genetic resources suitable for the smallholder production system. There is also an indirect loss as monetary resources are wasted that could be used later in suitable genetic improvements.

5.3 Impact of gene flow on biodiversity The influx of high performing genotypes into existing breeds has always been an important component in developing and improving breeds. In the history of all species investigated, gene flow has contributed significantly to diversity. In population genetics “migration” is an important source of genetic variability. The growing global need for improved animal productivity has led to an increased demand for improved genotypes. With biotechnology and global transport developed since the mid 20th century, foreign genetic resources are readily available across the globe. Current gene flow is dominated by source countries with advanced breeding structures, mainly developed countries, so that gene flow is north to north and north to south. At the same time, animals from developed countries increasingly belong to a small number of genetically narrow, high performance breeds with similar breeding goals and hybrids developed over the last two centuries and strongly influenced by controlled research- funded breeding programmes. They have been selected for high yields and require standardised conditions and high inputs to exploit their genetic potential and homogeneity within and between breeds has increased. The international exchange of genetic material from a decreasing number of sires increasingly loses genetic variation (Young and Seykora, 1996) with global impacts for developed and developing countries. The global gene flow of a relatively few breeds is significant for all species under study. Biodiversity is now recognised as relevant to global politics, and has been too long neglected in research and its economic relevance underestimated (Winnacker, 2005). The impact of this gene transfer on biodiversity depends mainly on the breed’s suitability to its new production environments and can be measured in changes to the national population composition. If they are suitable and the new breeds add to existing genotypes, then biodiversity increases. If new genotypes replace existing breeds, biodiversity is lost. Examples for both situations are numerous for all species and in general they are examples of national economic development. Whether the improved breeds co-exist with or replace the native breeds after gene flow depends on various factors. One factor is the production system that the introduced genotype is meant for. In developing countries, where extensive smallholder production systems prevail, dilution or replacement, mainly through crossbreeding, are likely where the new genotypes are suitable, are available, and are adopted by the local breeders. Then, developing smallholder livelihoods can conflict with improving biodiversity and it may not be adequate to require poor farmers in developing countries to bear the costs for conservation of global biodiversity. The more commercial and concentrated the sector is the greater the threat of losing biodiversity, a general rule which can be observed in other sectors likewise. Semyonov and Selkin (1989) for example summarised in their study on sheep breeds in Russia that in many decades of effort new sheep breeds were developed by crossbreeding local coarse- wooled sheep with fine-wooled and semi-fine-wooled sheep. However the local Voloshian, Mikhnov, Bozakh, Karabakh, Shirvan, and Darvaz breeds had almost disappeared, and the Fat-tailed Finewool, Georgian Semifinewool Fat-Tailed and Kuchugury were reduced to 90 ANALYSIS OF DRIVING FACTORS AND IMPACTS small populations. The country reports give other specific examples where local stock was replaced and breeds became extinct. For Africa, Eastern and Southern Europe, Western and Northern Europe, North America, and Western Asia however the threat of sheep breed extinction is considered low while in Latin America and Southern and Eastern Asia it is medium. The risk status of the world’s sheep breeds recorded up to end of 1999 as published by Scherf (2000) in the World Watch List for Domestic Animal Diversity shows that from a total of 1495 breeds, 181 (12%) were extinct while 656 (44%) were not at risk. Compared to other species, the total number of breeds, the number of extinct and not at risk breeds, indicate a better situation for sheep breeds than for cattle and pig breeds. Only goats had lower extinction rates (3%). However, the situation for sheep is deteriorating over time as extinction rates compared to figures from 1995 had more than doubled and the percentage of breeds not at risk decreased. However, the high numbers of breeds with unknown status and the still increasing number of recorded breeds shows these figures are uncertain. According to data collected by the FAO, 18 % of the 740 farm animal breeds that were recorded as extinct were breeds from the South, although the percentage may be underestimated as data collection started later and is more difficult in developing countries (Mathias and Mundy, 2005). However, among the breeds at risk, including the status endangered and critical, 60% are from the South and this proportion is expected to increase. The figures differ greatly between sources and between species and their reliability is further weakened by the great number of breeds with unknown status. However, Mathias and Mundy (2005) conclude that if the risk factors “change of husbandry”, “expansion of large-scale intensive livestock production” or “people giving up herding or farming” are taken into account, then the South would be the “hotspot” of breed loss of the 21st century. Niche market production using local breeds to serve specific local consumer preferences may counter these losses. Careful production system analysis is therefore demanded to evaluate the impact of gene flow on biodiversity, as well as to determine whether conserving local breeds is in the interest of local farmers. In Europe we can see a re- orientation towards the use of local breeds, due to new utilities based on eco-tourism and conserving natural environments. If the introduced breeds are not suitable, as already mentioned above, the impact on biodiversity is theoretically limited. However, if the introduction is promoted or even subsidised through development projects or governments, or crossbreeding is indiscriminate over a long time and in considerable quantities, existing breeds will be affected. The effect is generally negative as the native and adapted breeds are diluted. However, over a long adaptation time, these new genotypes can also contribute to biodiversity by developing new breeds or traits. In Latin America for example, introducing unsuitable wool sheep breeds from Spain led eventually to the well adapted Chiapas sheep, now considered “native”. Apart from the main streams of gene flow of high-yielding breeds from developed countries, gene flow takes place from all countries which possess unique genotypes. High- yielding breeds formed in developed countries depended on foreign genetic material. For example the European and North American commercial pig breeds were heavily influenced by genetic material from Asia. Research and commercial interests focus at such genotypes of current and future importance. ANALYSIS OF DRIVING FACTORS AND IMPACTS 91

Global transfer of these genotypes generally contributes to biodiversity, particularly if these genotypes are threatened in their native countries. Horst and Reh (1999) give examples where some “old” European breeds, under threat in their native country, are used for crossbreeding in the tropics. By making use of their typical adaptability to marginal production environments, valuable genetic resources from developed countries can be conserved as well as developing economies in developing countries by genetically improving local breeds. Niche market production using foreign breeds with specific attributes to serve local market preferences are also examples of the benefits of biodiversity by transferring unique genotypes. The utilisation of major genes is in this context of particular interest (Horst and Reh, 1999). Shrestha (2005) points at developing composite sheep populations from a combination of endangered and established breeds which may result in financial benefits and simultaneously conserve domestic animal diversity of the gene pool, although not of breeds. However, these are singular cases and are not meant to suggest that an overall model solution is to align the global interest to conserve biodiversity with the interests of local farmers to quickly increase production. In general, if global gene flows that reduce biodiversity are to be counteracted, creating incentives for smallholders in developing countries such as paying to conserve global biodiversity through local genetic resources may be more promising than imposing restrictions.

5.4 Limitations of the study The most limiting factor of the study was the information available on current gene flow. Countries with highly organised breeding structures were expected to have statistical data bases on breeding stock transfer and changes in national population composition. The United Nations Commodity Trade Statistics Database (COMTRADE) delivered unreliable results and was therefore dropped as a source. Europe had available statistics in the form of the Eurostat statistic database. Also these records had limited information. First of all, the records did not include breed details so that the influence of single breeds could not be interpreted. It was also unclear whether these records reflected true volumes of animal transfer, for example whether transfers for development projects were included. Furthermore, only the exchange of purebreeding stock was recorded. Purebreeding statistics are however not important where hybrids play a vital role, particularly in pig breeding. Additionally, with increasing commercialisation, export and import statistics are less able to depict gene flow and company figures would be more suitable. However international breeding companies tend to restrict access to this information at least to up- to-date figures. In commercial breeding, transfer figures underestimate the gene flow when a few animals are multiplied in national centres, is the case in commercial pig breeding. Apart from cattle, data on the exchange of other breeding products, such as semen and embryos, were even more difficult to obtain and in general lacking. Additionally, there may be big gaps between what has been transferred and what has been successfully utilised. Another major methodological restriction is that the availability and quality of information on transfers varies considerably between countries and extrapolating from one well documented situation to cover the many information gaps is not possible. For example, to gain information of the desired quality covering at least examples from the different regions of the world, six breeding organisations and 21 experts from 12 countries 92 ANALYSIS OF DRIVING FACTORS AND IMPACTS were contacted for the sheep global study. Only three contacts replied and no statistics were made available. Official national statistics were available via the FAO on the exchange of live animals but these were not exclusively for breeding. The country reports in some cases included necessary information. However, if country reports included data on animal transfers, the usual focus was on imports not exports. Exports could only be traced indirectly as imports into other countries. Therefore, an important source of recorded information on gene flow was generally left out. This was particularly bad for developing countries, where records of animal movements rarely existed and country reports, if they existed, did not contain the desired details. In general country reports differed considerably in quantity and quality and no standard format exists which would allow such information to be compared. Export data from developed countries could have closed that information gap. The latter limitation throws light on the lack of information about gene flow from developing countries, particularly South to South movements which would have been interesting. A major constraint is that the size and number of transfers does not generally predict its genetic or economic impact. Transactions are, with very exceptions, impossible to follow up. The exceptions, such as the Awassi case study, cannot be generalised. There were examples for all species where massive transfers had very little long-term effect, and where the transfer of a few animals had major impacts on global gene flow. These examples however only become evident retrospectively as changes to national population composition, making this indicator of gene flow an important tool in the study. It was used especially in the global study but relied mainly on the Country Reports which, as mentioned above, were limited in quantity and quality. Additionally, a major limitation of the country reports on changes in national population composition was put forward in the pig case study. According to more updated sources it appears that the Vietnamese country report did not realise the full influence of exotic genotypes on local genotypes, and only reported where both were directly interfering. So the impact of the exotic genotypes on the local populations was underestimated. With the increase of global mobility and biotechnology, gene flow has increased and has become more complicated and even more difficult to follow up. This puts severe limitations on the global studies. The Awassi case study showed in an illustrative example how this limitation could be overcome in particular, detailed case studies. But it also showed what effort it takes to collect the necessary information to assemble a nearly complete picture of gene flow for one particular breed from one particular regional source. The study as a whole therefore had to refrain from quantifying global gene flow. It was shown that gene flow is a very complex and dynamic process with many factors varying the possible impacts from case to case. The impact evaluation of global gene flow on economic development and biodiversity as treated in chapter 5.2 and 5.3 therefore make very general and qualitative conclusions and do not measure overall global gene flow. The study analyses selected aspects of gene flow and is meant to provide scientifically-founded information for decision makers in the current discussion on global gene flow and its impacts. It may help to understand the complexity of the situation and may prevent decision makers from implementing oversimplified programmes. The study is certainly not suitable to conclude this discussion. ANALYSIS OF DRIVING FACTORS AND IMPACTS 93

5.5 Need for further action The discussion on gene flow can only be fruitful if the impact of gene flow can be suitably evaluated. It has been stated above that global evaluation is doubtful and that global interpretations will be possible only if serious status and impact studies in each country add to the global picture. With the limitations of the data bases accessed in this study, further actions are needed to improve the information on current gene flow in a standardised comparable format for each species and country if a global valuation is intended. As improving data quality and quantity in developed as well as developing countries, particularly import and export figures, are not realistic in the near future the focus should be to derive reliable data on national population composition changes. This calls for frequent census studies at breed and genotype level. This approach, which for most countries is challenging enough, would avoid the time consuming tracing of complex gene flow routes and instead starts right at evaluating the genetic impact and can be more easily connected to collateral impacts on biodiversity and economic development for different groups of society. Many weaknesses could be minimised, such as the different values put on national genetic resources by different stakeholders. Case studies then need to provide additional information on the impact evaluation for very specific countries or regions as well as on the organisational structures and mechanisms influencing the impact. Each country, based on its evaluation of the national situation, can then make use of gene flow control mechanisms to improve the situation. One has to be aware that indigenous genetic resources are a source of adaptability to specific changes in the economic and natural environment. Conserving these resources secures the national as well as global adaptability to changing future animal production conditions, and secures the livelihoods of animal holders who for the time being depend on these resources. Future animal production has to maintain developing countries’ biodiversity and reduce poverty. Concepts which serve both by securing or even improving smallholder livelihoods in marginal areas using their indigenous genetic resources should be heavily weighted. These concepts need to identify and evaluate the genetic distance between breeds as well as the genetic potential, performance and other uses and non-use values in different production systems. What is more, biodiversity impacts must be taken into account when conditions of gene flow are evaluated. Exotic breeds are currently introduced for various reasons. If the genetic potential of the local breed is low, economic development calls for introducing suitable higher performing breeds. Suitability requires reasonable selection of imported breeds, comparative performance evaluation of different pure- and crossbred genotypes, and evaluation of natural, economic, social and cultural framework conditions, before selecting a genotype and breeding system. Conserving national genetic resources by crossbreeding can be considered as well as using foreign breeds which are under threat in their country of origin. Such approaches should preferably be carried out on-farm closely involved with the livestock keepers. However, if conserving indigenous genetic resources conflicts with economic development and poverty reduction, conservation cannot be carried out at the expenses of local livestock keepers. Conservation then has to be treated separately but alongside economic development, supported by appropriate funding. Herders and farmers must be closely involved in decision making, and appropriate system analysis must be developed for smallholder production systems in developing countries 94 ANALYSIS OF DRIVING FACTORS AND IMPACTS

(Gibson and Pullin, 2005). Based on its central importance to reduce poverty and conserve biodiversity in marginal areas, community based management of animal genetic resources and indigenous knowledge has to be fostered. Each country has to be aware that unique genotypes are in global demand. Gene flow is too complicated to find routes after the event, in whatever quantity. Gene flow should therefore be considered case by case, and recording practices and access and benefit sharing agreements put in place in advance, rather than retrospectively attempting to measure and exploit the transfers. The apparent lack of prior informed consent is one of the major discussion points in the case of the Boran and Tuli genetic resources transfer, as detailed in the case study. Finally it has to be concluded that without local breeding organisations and without actively including smallholders in the breeding work success is impossible in the long run. Instead, international breeding companies are ready to do that job. As already observed in cattle and pig breeding, a two-level structure is emerging: modern and traditional. What is more, as the demand for meat and milk in developing countries is expected to double in the next two decades, a new “livestock revolution” is expected. In the long run, a rapid expansion of intensive production may put traditional production under increasing pressure (Mathias and Mundy, 2005). However, the importance of animal husbandry for the livelihoods of smallholders, particularly in marginal regions, cannot be overestimated. If regional structural changes do not allow these groups to find new incomes, then breeding should be organised and payments to conserve natural environments and farm animal diversity should be considered world wide. REFERENCES 95

6 REFERENCES

Afifi-Affat, K.A. 1998. Heifer-in-Trust: a model for sustainable livestock development. World Animal Review 91:13-20. Ashley, S., J. Gray, S. Holden, J. Tanner, N. Schlaepfer, M. Andrew, L. Stockley, and B. Taylor (2002): Poverty and Livestock Breed Diversity - the way forward for DFID, Draft consultation document, The IDL Group. Agbagha, F.M., F.U. Ezema, and B.C.O. Omeke. 2001. Studies of management effects on fertility of purebred and crossbred exotic gilts in two breeding farms at Nsukka, Nigeria. Nigerian Journal of Animal Production 28:20-25. Alba, J.D. 1978. Progress in the selection of the Latin American Dairy Criollo. World Animal Review 28:26-30. Allen, J. 2002. Country Report: Australia. ACIAR Proceedings Series Number 108: Development strategies for genetic evaluation for beef production in developing countries. 13-22. Arganosa, A.S., L.T. Trung, and E.M. Rigor. 1989. Lactation performance of imported Sahiwal-Holstein crosses in the Philippines under three management systems. Philippine Journal of Veterinary and Animal Science 15:1-8. Arora, A.L., R.C. Sharma, and B.K. Khan. 2002. Sustainable mutton production in sheep breeds of India. 7th World Congress on Genetic Applied to Livestock Production, 19-23August 2002, Montpellier, France. ASPS. 2002. Livestock policy briefs for Vietnam. MARD - DANIDA: Agriculture Sector Programme Support ASPS; Agrifood Consulting International ACI, Hanoi, Vietnam. Banos, G. and A. Sigurdsson. 1996. Application of Contemporary Methods for the Use of International Data in National Genetic Evaluations. Journal Dairy Sciences 79:1117- 1125. Benyoucef, M.T. 1992. The local sheep breeds of Algeria: Situation and perspectives. In: Galal E.S.E. and O. Gürsoy. 1992. Strategies for the development of fat-tail sheep in the Near East. Proceedings of the workshop organised by FAO, ICARDA, EAAP, CIHEAM and CUZF, 5-9 October 1992, Adana, Turkey, EAAP Publication No. 68. Blake, R.W. 2004. Dairy cattle performance in difficult environments. 7th World Bruna Conference, Fieragricola, Verona, Italy. 125-131. Blench, A. 1999. Traditional Livestock breeds: Geographical Distribution and Dynamics in Relation to the Ecology of West Africa. London, Overseas Development Institute Portland House. Working Paper 122. Blench, A. and K.C. MacDonald. 2000. The origins and development of African livestock: archaeology, genetics, linguistics and ethnography. UCL Press, 2000. - XX, 546 S. Boadle, R., M. Coombe, and J. Croaker. 1997. Pedigree indexing of imported breeder cattle in the Philippines. Asian-Australasian Journal of Animal Sciences 10:614-620. 96 REFERENCES

Boer Goat Breeders’ Association of Australia. 2004. Australia’s Boer goats: Meating the market. http://boergoat.une.edu.au/. (01.12.2005) Bökönyi, S. 1974. History of domestic mammals in Central and Eastern Europe. Akademiai Kiado, Budapest. 597 pp. Bondoc, O.L., C.M. Mercado, N.C. Vera-Cruz, R.N. Palou, and J.S.J. Server. 1997. Crossbreeding dual-purpose cattle for beef production in tropical regions. Meat science 47:177-185. Brahman Cattle Breeders’ Society South Africa. 2004. Evaluation of the productivity of the Brahman, with other breeds and their crosses in East and Southern Africa. http://www.brahman.co.za/main.htm. (01.12.2005) Brahmia, A. and M. Ben Dhia. 1990. The problem of the genetic improvement of cattle in Tunisia. Amelioration genetique de bovins sous climat sud mediterraneen. EAAP Publication, No. 47:12-24. Brascamp, E.W. 1998. The development of animal breeding enterprises in the next decades: reasons to change the structure of breeding programmes. Acta Agriculturae Scandinavica Supplementum 28:61-67. Chagunda, M.G.G. 2000. Genetic Evaluation of the Performance of Holstein Frisian Cattle on large-scale dairy farms in Malawi. Cuvillier Verlag, Göttingen. Chander, M., S. Kumar, and P.C. Harbola. 2000. Dynamics of goat husbandry in Indian Himalayan region. 7th International Conference on Goat, France, 15-21 May 2000, 334-336. Cossee, O. 2003. Emergency Farm Reconstruction Project Kosovo - Analysis of the Final Survey. FAO, Rome, Italy. Clutton-Brock, J. 1981. Domesticated animals from early times. Heinemann/British Museum Natural History. 208 pp. Cunningham, E.P. 1991. Breeding programmes for improved dairy production in tropical climates. Animal husbandry in warm climates. EAAP Puplication 55:39-47. Cunningham, E.P. and O. Syrstad. 1987. Crossbreeding Bos indicus and Bos taurus for milk production in the tropics. FAO Animal Production and Health Paper 68. FAO, Rome, Italy. CSIROnline. 2004. New beef quality gene marker identified. http://www.csiro.au/index.asp?type=mediaRelease&id=Prtenderbeef3. (01.04.2004) Cunningham, E.P. 1991. Breeding programmes for improved dairy production in tropical climates. Animal husbandry in warm climates. EAAP Puplication 55:39-47. Cunningham, E.P. and O. Syrstad. 1987. Crossbreeding Bos indicus and Bos taurus for milk production in the tropics. FAO Animal Production and Health Paper 68. FAO, Rome, Italy. Cutrer, G. 1995. Boer goats for beginners. Ranch and Rural Living Magazine 77(2). DAD-IS. 2004. FAO information system for the Global Strategy for the Management of Farm Animal Genetic Resources. http://www.fao.org/dad-is. (01.12.2004) Daly, J.J. 1981. Beef cattle breeds. Queensland Agricultural Journal 107:28-50. REFERENCES 97

Davis H.D., S.M. Galloway, I.K. Ross, S.M. Gregan, J. Ward, B.V. Nimbkar, P.M. Ghalsasi, C. Nimbkar, G.D. Gray, Subandriyo, I. Inounu, B. Tiesnamurti, E. Martyniuk, E. Eythorsdottir, P. Mulsant, F. Lecerf, J.P. Hanrahan, G.E. Bradford, and T. Wilson. 2002. DNA tests in prolific sheep from eight countries provide new evidence on origin of the Booroola (FecB) mutation. Biology of Reproduction 66:1869-1874. De Haan, N.C., K. Tibbo, E. Kinsey, and C. Valdivia. 2000. Social dimensions of goat projects in East Africa. 7th International Conference on Goat, France, 15-21 May 2000, 507-513. Drucker, A., E. Bergeron, U. Lemke, L.T. Thuy, and A. Valle Zárate. 2005. Identification and quantification of subsidies relevant to the production of local and imported pig breeds in Vietnam. Tropical Animal Health and Production Submitted. El Aich, A., S. Landau, M. Napoleone, and A. Bourbouze. 1995. Goat production systems in the Mediterranean: comparative study. In: A. El Aich, S. Landau, A. Bourbouze, R. Rubino, and P. Morand-Fehr (eds.) Goat production systems in the Mediterranean: Results of an FAO (REU/RNE)/CIHEAM working group on goat production. EAAP Publication No. 71. pp. 222-237. Wageningen Pers, Wageningen, The Netherlands. Epstein, H. and M. Bichard. 1984. Pig. In: Mason, I.L. (ed.) Evolution of domesticated animals. Longman, London and New York. Fahmy, M.H. 1996. Prolific sheep. CAB International, Wallingford, UK. FAO. 2001. Guidelines for the Development of Country Reports. FAO, Rome, Italy. http://www.fao.org/dad-is. (01.06.2004) FAO. 2004a. Global Livestock Production and Health Atlas (GLiPHA). 1[1]. 2004. FAO, Rome, Italy. FAO. 2004b. The legal framework for the management of animal genetic resources. Commisson on genetic ressources for food and agriculture. FAO - Background study paper No. 24. FAO, Rome, Italy. FAO-aktuell. 2000. Nachrichten über Welternährung und Weltlandwirtschaft. 30. Jahrgang, Nr. 46-47/2000. FAO, Rome, Italy. Felius, M. 1995. Cattle breeds: an encyclopedia. Doetinchem: Misset. Fitzhugh H.A. and G.E. Bradford. 1983 Hair sheep of Western Africa and the Americas - A genetic resource for the tropics, Westview Press, Boulder, Colorado, USA. Gall, C. 1999. Förderung der Tierproduktion. In: Horst, P. and Reh I. 1999. Tierzucht in den Tropen und Subtropen. Handbuch der Landwirtschaft und Ernährung in den Entwicklungsländern, Ulmer Verlag, Stuttgart, Germany. Gall, C. 2005. Personal communication. Gamble, H.R. and A.M. Zajac. 1992. Resistance of St. Croix lambs to Haemonchus contortus in experimentally and acquired infections. Veterinary Parasitology 41(3- 4):211-225. Gatenby, R.M. 1986. Sheep production in the tropics and subtropics; Longman, London and New York. 98 REFERENCES

Gatenby, R.M., M. Doloksaribu, and E. Romjali. 1994. The potential of hair sheep for the humid tropics of Southeast Asia. 7th Asian-Australian Association for Animal Production Societies (AAAP) Congress, 11-16 July 1994, Bali, Indonesia. 25-38. Gatenby, R.M., G.E. Bradford, M. Doloksaribur, E. Romjali, A.D. Pitono, and H. Sakuhl. 1997a. Comparison of Sumatra sheep and three hair sheep crossbreds. I. Growth, mortality and wool cover of F1-lamb. Small Ruminant Research 25:1-7. Gatenby, R.M., M. Doloksaribu, G.E. Bradford, E. Romjali, A. Baubara, and I. Mirza. 1997b. Comparison of Sumatra sheep and three hair sheep corssbreds. II: Reproductive performance of F1-ewes. Small Ruminant Research 25:161-167. Gibson, J.P. and R.S.V. Pullin. 2005. Conservation of livestock and fish genetic resources. Science Council Secretariat, FAO, Rome, Italy. Glimp, H. 1995. Meat goat production and marketing. Journal of Animal Science 73:291- 295. Goossens, B., S. Osaer, S. Kora, J. Jaitner, and M. Ndao. 1997. The interaction of Trypanosoma congolense and Haemonchus contortus in Djallonke sheep. International Journal of Parasitology 27(12):1579-1584. Goossens, B., S. Osaer, M. Ndao, J. van Winghem. 1999. The susceptibility of Djallonke and Djallonke-Sahelian crossbred sheep to Trypanosoma congolense and helminth infection under different diet levels. Veterinary Parasitology 85 (1):25-41. Grupp, T. 2003. Spreading the best Fleckvieh genetics to the world. Fleckvieh World 14- 17. Haring, F. 1984. Schafzucht. Ulmer Verlag, Stuttgart, Germany. Heifer International. 2004a. Multipurpose Cattle Production in Southwestern Poland, Phase II. http://www.heifer.org/Our_Work/Our_Projects/view_project.cfm?project_id= 2359&keyword=Simmental&country=®ion=&issue=. (19.07.2004) Heifer International. 2004b. Not Just Any Cow Will Do. http://www.heifer.org/Our_Work/ Success_Stories/In_Central_Eastern_Europe/Ukrainian_Cattle.shtml. (19.07.2004) Heifer International. 2004c. Simmental Cattle for Farmers in Dragashi Region. http://www.heifer.org/site/apps/nl/content3.asp?c=edJRKQNiFiG&b=718993&ct=91 3929. (19.07.2004) Heifer International. 2004d. Turpan Beef and Dairy Cattle Project. http://www.heifer.org/ Our_Work/Our_Projects/view_project.cfm?project_id=2435&keyword=Simmental& country=®ion=&issue=. (19.07.2004) Henson, E.L. 1992. In situ conservation of livestock and poultry. FAO Animal Production and Health Paper 99. FAO, Rome, Italy. Hernandez-Ceron, J., C.C.Jr. Chase, and P.J. Hansen. 2004. Differences in heat tolerance between preimplantation embryos from Brahman, Romosinuano, and Angus breeds. Journal of Dairy Science 53-58. Horst, P. 1999a. Stand der Nutztierproduktion in der Welt. In: Horst, P. and Reh I. 1999. Tierzucht in den Tropen und Subtropen. Handbuch der Landwirtschaft und Ernährung in den Entwicklungsländern, Ulmer Verlag, Stuttgart, Germany. REFERENCES 99

Horst, P. 1999b. Züchtung. In: Horst, P. and Reh I. 1999. Tierzucht in den Tropen und Subtropen. Handbuch der Landwirtschaft und Ernährung in den Entwicklungsländern, Ulmer Verlag, Stuttgart, Germany. Horst, P. and Reh I. 1999. Tierzucht in den Tropen und Subtropen. Handbuch der Landwirtschaft und Ernährung in den Entwicklungsländern, Ulmer Verlag, Stuttgart, Germany. IFAD. 1984. Generation and Transfer of and Seed Production. http://www.ifad.org/lrkm/region/pl/gt_154.htm#garea. (13.08.2004) IFAD. 1988. Smallholder Livestock Development Project. http://www.ifad.org/lrkm/ region/ pi/ ph_84.htm. (13.08.2004) IFAD. 2004a. Inputs and Infrastructure - Restocking. http://www.ifad.org/lrkm/range/ restock.htm. (13.08.2004) IFAD. 2004b. IFAD Livestock Development Project - Samoa. www.ifad.org/lrkm/region/ pi/ws_75.htm. (13.08.2004) Ikiror, D. 2001. Heat Stress in Dairy Cattle in Kenya. Tropical Animal Production. Cuvillier Verlag, Göttingen. Iniguez, L., 1998. Community breeding programs for small ruminants in the Andean region. 6th World Congress of Genetics Applied to Livestock Production, 12-16 January 1998, Armidale, Autralia. 25:249. Ishaq, S.M., M.A. Khan, and S.K. Shah. 1981. Experimental crossbreeding of Sahiwal cows for improved production performance. Pakistan Veterinary Journal 1:154-160. Israeli Cattle Breeder’s Association. 2004. Breeding. http://www.icba-israel.com/icba- breeding.htm (11.07.2005) Jabbar, M.A. and M.L. Diedhiou. 2003. Does breed matter to cattle farmers and buyers? Evidence from West Africa. Ecological Economics 45:461-472. Jeyaruban, M.G., G. Giritharan, P.A.L. Senadeera, and S. Attanayake. 1997. Meat goat production in the Dry Zone of Sri Lanka. In: Proc. of the 14th Annual Scientific Sessions of Sri Lanka Veterinary Association. BMICH, Colombo, Sri Lanka. Johnson, D.W., A. Tber, and A.T. Benkirane. 1989. Importation au Maroc de bovins Holstein américains: considérations sanitaires. Maghreb Vétérinaire 4:11-14. Kahi, A.K., W. Thorpe, and C.F. Gall. 1999. Crossbreeding for dairy production in Kenya: Parameter estimates for defining optimal crossbreeding systems. Tropentag, Berlin/Germany, 14.10.-15.10. 1999. Kahi, A.K., G. Nitter, W. Thorpe, and C.F. Gall. 2000. Crossbreeding for dairy production in the lowland tropics of Kenya: II. Prediction of performance of alternative crossbreeding strategies. Livestock Production Science 63:55-63. Katpatal, B.G. 1977. Dairy cattle crossbreeding in India. World Animal Review 23:2-9. Kimenye, D. and W.S. Russell. 1975. Comparison of Ayrshire x Sahiwal cows with high grade Ayrshire in Kenya. East African Agricultural and Forestry Journal 40:416-421. 100 REFERENCES

Kosgey, I.S., J.H.J. van der Werf, B.P. Kinghorn, J.A.M. van Arendonk and R.L. Baker. 2002. 7th World Congress on Genetic Applied to Livestock Production, 19-23August 2002, Montpellier, France. Letenneur, L. 1978. Crossbreeding N’Dama and in Ivory Coast. World Animal Review 26:36-42. Loculan, D. 2002. Genetic evaluation of beef production in the Philippines. Development strategies for genetic evaluation for beef production in developing countries 72-80. Ludwig, A. 1997. Abstammung der europäischen Hausschafe und Phylogenie der eurasischen Wildschafe. Dissertation, Humboldt-Universität, Berlin, Germany. Machen, R.V. 1997. Boer goats - introduction and impact in North America. Journal of Animal Science 75:139. Madalena, F.E. 1977. Crossbreeding systems for beef production in Latin America. World Animal Review 22:27-33. Madalena, F.E., A.M. Lemos, R.L. Teodoro, R.T. Barbosa, and J.B.N. Monteiro. 1990. Dairy Production and Reproduction in Holstein-Friesian and Guzera Crosses. Journal Dairy Sciences 73:1872-1866. Maier, J., S. Gura, I. Köhler-Rollefson, E. Mathias, and S. Anderson. 2002. Livestock Diversity: Keepers’ Rights, Shared Benefits and Pro-Poor Policies. Bonn, 2002, German NGO Forum on Environment and Development. Mandibaya, W., C. Mutisi, and H. Hamudikuwanda. 1999. Cow production from upgrading Brahman to Nelore and Guzerat. Journal of Animal Breeding and Genetics 116:75-86. Mason, I.L. 1981. Wild goats and their domestication. In: Gall, C. (ed.) Goat Production. pp. 35-53. Academic Press, London. Mathias, E. and P. Mundy. 2005. Herd movements. The exchange of livestock breeds and genes between North and South. League for Pastoral Peoples and Endogenous Livestock Development, Ober-Ramstadt, Germany. Mavena, K. 1987. High degree of tolerance to trypanomiasis of dwarf Djallonke sheep and goats in the Southern Guinea area of Togo in comparison with trypanosoma tolerant cattle. Revue d’Elevage et de Medecine Veterinaire des Pays Tropicaux 40(1):55-8 McDowell, R.E. 1985. Crossbreeding in tropical areas with emphasis on milk, health, and fitness. Journal of Dairy Science 68:2418-2435. McDowell, R.E. 1992. Dairy production in developing countries. Proceedings of the New Zealand Society of Animal Production 52:3-5. McDowell, R.E., J.C. Wilk, and C.W. Talbott. 1996. Economic viability of crosses of Bos taurus and Bos indicus for dairying in warm climates. Journal Dairy Sciences 79:1292-1303. Meyn, K. 2005. Personal communication. Meyn, K. and J.V. Wilkins. 1974. Breeding for milk in Kenya, with particular reference to the Sahiwal stud. World Animal Review 11:24-30. REFERENCES 101

Mpofu, N. and J.E.O. Rege. 2002. Monitoring of Sahiwal and Friesian cattle genetic improvement programmes in Kenya. In: ILRI-SLU 2003. Animal Genetics Training Resource CD-ROM. Version 1. Mwai O. and B. Malmfors (eds.). ILRI (International Livestock Research Institute), Nairobi, Kenya and SLU (Swedish University of Agricultural Sciences), Uppsala, Sweden. Mudgal, V.D. and C.L. Arora. 1994. Frieswal project: present status and expectations for the future. World Animal Review 79:46-50. Mueller, J.P., E.R. Flores, and G.A. Gutierrez. 2002. Experiences with a large scale sheep genetic improvement project in the Peruvian Highlands. 7th World Congress on Genetic Applied to Livestock Production, 19-23August 2002, Montpellier, France. Mueller, J. 2004. Personal communication. Mugambi, J.M., R.K. Bain, S.W. Wanyangu, M.A. Ihiga, J.L. Duncan, M. Murray, M.J. Steak. 1997. Resistance of four sheep breeds to natural and subsequent artificial Haemonchus contortus infection. Veterinary Parasitology 69(3-4):265-273. Murugaiyah, M., P. Ramakrishnan, A.R. Sheikh Omar, C.H. Knight, and C.J. Wilde. 2001. Lactation failure in crossbred Sahiwal Friesian cattle. Journal of Dairy Research 68:165-174. Nakimbugwe, H. 2005. Open Nucleus Dairy Cattle Breeding Programme in the Lake Victoria Crescent Region of Uganda. Dissertation. Institut für Nutztierwissenschaften, Department für Nachhaltige Agrarsysteme, Universität für Bodenkultur Wien. Narrod, C.A. and K.O. Fuglie. 2000. Private investment in livestock breeding with implications for public research policy. : An International Journal 16:457-470. Nassuns-Musoke, M. 2002. Environmental Constraints depress the Annual Milk Yield of Frisian Cows on Small Dairy Farms around Kampala. Cuvillier Verlag, Göttingen. Nimbkar, C., P.M. Ghalsasi, S.W. Walkden-Brawn, and L.P. Kahn, 2002. for the genetic improvement of Deccani sheep of Maharashtra, India. 7th World Congress on Genetic Applied to Livestock Production, 19-23August 2002, Montpellier, France. Nozawa, K. 1983. Domestication and history of goats. In: Neimann-Sorense, A. and D.E. Tribe (eds.) World animal science. Elsevier, Amsterdam, The Netherlands. Nsubuga, H.S.K. 1996. Small ruminants: Goats and sheep in Uganda. In: S.H.B. Lebbie and E. Kagwini (eds.) Small Ruminant Research and Development in Africa. Proceedings of the Third Biennial Conference of the African Small Ruminant Research Network, UICC, Kampala, Uganda, 5-9 December 1994. ILRI, Nairobi, Kenya. Padbury, T.A.D. 2002. The Australian Merino industry (1788-2002). 6th Merino World Conference, Budapest, Hungary. Payne, W.J.A. and R.T. Wilson. 1999. An introduction to animal husbandry in the tropics. 5th Edition. Blackwell Science, Oxford. 102 REFERENCES

Payne, W.J.A. and J. Hodges. 1997. Tropical cattle: origins, breeds, and breeding policies. 1st Edition. Blackwell Science, Oxford. Pelant, R.K., B. Chandra, J.B. Pu, M. Lohani, N. Suknaphasawat, and G. Xu. 1999. Small ruminants in development: the Heifer Project International experience in Asia. Small Ruminant Research 34:249-257. Pelant, R.K. 2001 Goats for economic development. Proc. of the 2001 Conference on Boer goats in China, 21-24 October, Guizhou, 33-41. Philipsson, J. 2002. Sustainability of dairy cattle breeding systems utilising artificial insemination in less developed countries - examples of problems and prospects. In: ILRI-SLU 2003. Animal Genetics Training Resource CD-ROM. Version 1. Mwai O. and B. Malmfors (eds.). ILRI (International Livestock Research Institute), Nairobi, Kenya and SLU (Swedish University of Agricultural Sciences), Uppsala, Sweden. Polge, C. 1985. How does embryo manipulation fit into present and future pig reproduction? Journal of Reproduction and Fertility Supplement 33:93-100. Ponzoni, R.W. 1992a. A global review of the genetic resources of sheep and goat breeds. In: Hodges, J. (ed.) The management of global animal genetic resources. FAO Animal Production and Health Paper 104. FAO, Rome, Italy. Ponzoni, R.W. 1992b. Genetic improvement of hair sheep in the tropics. FAO Animal Production and Health Paper 101. FAO, Rome, Italy. Porter, V. 1996. Goats of the world. Farming Press, Ipswich, UK. Preston, J.M. and E.W. Allonby. 1979. The influence of breed on the susceptibility of sheep of Haemonchus contortus infection in Kenya. Research of Veterinary Sciences, 26(2):134-139 Rajapaksha, W.R., M.G. Jeyaruban, and C.M.B. Dematawewa. 2001 Effect of cross breeding on weight gains of crossbred (local x Boer) goats managed under small holder extensive management system in the dry zone of Sri Lanka. International Goat Association, 215-218. Reed, C.A. 1984. The beginnings of animal domestication. In: Mason, I.L. (ed.) Evolution of domesticated animals. pp. 1-6. Longman, London and New York. Rege, J.E.O. 1998. Utilization of exotic germplasm for milk production in the Tropics. Procedings of the 6th World Congress on Genetics Applied to Livestock Production, 12-16 January 1998, Armidale, Autralia. 25:193-200. Röhrs, M. 1994. Entwicklung der Haustiere. In: Kräußlich, H. and G. Brem (eds.) Tierzüchtungslehre. Ulmer Verlag, Stuttgart, Germany. Sansthan, L.P.-P. and I. Köhler-Rohlefson. 2005. Indigenous breeds, local communities. Documenting animal breeds and breeding from a community perspective. Lokhit Pashu-Palak Sansthan (LPPS), www.lpps.org, Mudra, Sadri, Rajasthan, India. Schäfer, C. 1998. The influence of crossbreeding Cameroon hair sheep with West Malaysian wool sheep on coat characteristics. Dissertation, Humboldt-Universität, Berlin, Germany. REFERENCES 103

Scherf, B. 2000. World Watch List for Domestic Animal Diversity. 3rd Edition. FAO publication. http://dad.fao.org/en/refer/library/wwl/wwl3.pdf. (01.06.2005) Schoeman, S.J. 1996. Characterization of beef cattle breeds by virtue of their performances in the National Beef Cattle Performance and Progeny Testing Scheme. South African Journal of Animal Science 26:15-19. Semyonov, S.I. and I.I. Selkin. 1989. Sheep. In: Animal Genetic Resources of the USSR. FAO Animal Production and Health Paper No. 65. N.G. Dmitriev and L.K. Ernst (eds.), FAO, Rome. Shelton, M. and E.A.P. Figueiredo. 1990. Hair sheep production in tropical and subtropical regions. Small Ruminant Collaborative Research Support Program, Davis, USA. Shrestha, J.N.B. 2005. Conserving domestic animal diversity among composite populations. Small Ruminant Research 56: 3-20. Simm, G. 1998. Genetic Improvement of Cattle and Sheep. Farming Press, Ipswich, UK. Smith, C. and G. Banos. 1991. Selection within and across populations in livestock improvement. Journal of Animal Science 69:2387-2394. Smith, K. 2004. South African vs South African. Boer Goat and Meat Online Magazine 19[33]. Sonn, H. 1985. Deutsches Fleckvieh im Ausland-Ein wichtiger Beitrag zur Entwicklung der Welt-Fleckviehzucht. Tierzüchter 37:101-102. Syed, M., A. Rehman, M. Farooq, and M. Ali. 1996. Production performance of Sahiwal and various grades of cross bred cattle. Sarhad Journal of Agriculture 12:601-610. Syrstad, O. 1991. The role and mechanisms of genetic improvement in production systems constrained by nutritional and environmental factors. FAO Animal Production and Health Paper 86:48-55. Thorpe, W., C.A. Morris, and P. Kang’Ethe. 1994. Crossbreeding of Ayrshire, Brown Swiss, and Sahiwal cattle for annual and lifetime milk yield in the lowland tropics of Kenya. Journal Dairy Sciences 77:2415-2427. Trail, J.C.M. and K.E. Gregory. 1981a. Sahiwal cattle in Africa - how they can help improve milk and beef production. Span. Prog. Agric. 24:28-30. Trail, J.C.M. and K.E. Gregory. 1981b. Sahiwal Cattle: An Evaluation of their Potential Contribution to Milk and Beef Production in Africa. ILCA Monograph [3]. International Livestock Centre for Africa. Vaccaro, R. 1983. Survival of imported Holstein and Friesian cattle and their locally born progeny in Venezuela. Tropical Animal Production 8:87-98. Valle Zaráte, A. 1996. Breeding strategies for marginal regions in the tropics and subtropics. Animal Research and Development 43/44:99-118 Wan Hassan, W.E., R.H. Phipps, and E. Owen. 1989. Development of smallholder dairy units in Malaysia. Tropical Animal Health and Production 21:175-182. Winnacker, E.-L. 2005. Bericht des Präsidenten der Deutschen Forschungsgemeinschaft anlässlich der Festveranstaltung am 6. Juli 2005 in Berlin im Rahmen der Jahresversammlung der DFG. http://www.dfg.de/aktuelles_presse/reden_ 104 REFERENCES

stellungnahmen/2005/download/jahresversammlung_festrede_winnacker_05.pdf. (01.02.2006) World Bank. 2003. Implementation completion report on a trust fund grant in the amount of $US18.4 Million to the United Nations interim administration mission in Kosovo for the benefit of Kosovo, Serbia and Montenegro (Kosovo) for an emergency farm recontruction project. 26932. http://www-wds.worldbank.org/servlet/WDSContent Server/WDSP/IB/2004/01/08/000090341_20040108095652/Rendered/PDF/269320 YF.pdf. (11.07.2005) World Simmental Fleckvieh Federation. 2002 Minutes of the Council Meeting of the World Simmental-Fleckvieh Federation. 19th and 23rd August 2002, Sheraton Hotel, Pretoria, South Africa. http://www.worldsimmentalfed.com/. (01.05.2005) Yazwinski, T.A., L. Goode, D.J. Moncol, G.W. Morgan, and A.C. Linnerud. 1980. Haemonchus contortus resistance in straightbred and crossbred Barbados Blackbelly sheep. Journal of Animal Science 51(2):279-284. Young, C.W. and A.J. Seykora. 1996. Estimates of and Relationship Among Registered Holstein Females in the United States. Journal Dairy Sciences 79:502- 505. Zeuner, F.E. 1963. A history of domesticated animals. Hutchinson, London. Zhong, X. 1976. Etude sur l’origine de la domestication du porc en Chine basee sur des ossements et des statuttes mis a jour a Hemudu. Vestiges culturels 8:24-26. Zhimin, A. 1991. Radiocarbon dating and prehistoric archaeology of China. World Archaeology. 23(2):193-200. Zwald, N.R., K.A. Weigel, W.F. Fikse, and R. Rekaya. 2003. Identification of factors that cause genotype by environment interaction between herds of Holstein cattle in seventeen countries. Journal of Dairy Science 86:1009-1018. CONTACT ADDRESSES 105

7 CONTACT ADDRESSES

7.1 Authors M.Sc. Erika Alandia Robles, Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart, Germany; Phone: +49 (711) 459 3006, Fax: +49 (711) 459 3290 E-mail: [email protected] Prof. em. Dr. Christian Gall, Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart, Germany; Phone: +49 (711) 459 3170, Fax: +49 (711) 459 3290 E-mail: [email protected] Dr. Elisha Gootwine, Institute of Animal Science, A.R.O., The Volcani Center, P.O. Box 6, Bet Dagan 50250 Israel; Phone: +97 (239) 683 752, Fax: +97 (239) 603 678 E-mail: [email protected] Martin Hill, Edinburgh Phone: +44 7901 55 24 66, http://www.martinhill.me.uk E-mail: [email protected] Dr. Sabine Homann, International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), POB 776, Bulawayo, Zimbabwe; Phone: +263 (11) 623 967, Fax: +263 (83) 8253/8307 E-mail: [email protected] Dr. Christian G. Hülsebusch, Executive Manager, DITSL GmbH Witzenhausen at the University of Kassel, Steinstrasse 19 , 37213 Witzenhausen, Germany; Phone: +49 (0) 5542 607 29, Fax: +49 (0) 5542 607 39 E-mail: [email protected]; [email protected] M.Sc. Le Thi Thanh Huyen, Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart, Germany, Phone: +49 (711) 459 3006, Fax: +49 (711) 459 3290 E-mail: [email protected] Dipl. Ing. agr. Ute Lemke, Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart, Germany, Phone: +49 (711) 459 3294, Fax: +49 (711) 459 3290 E-mail: [email protected] M.Sc. Jacobus Hendrik Maritz, Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart, Germany; Phone: +49 (711) 459 3172, Fax: +49 (711) 459 3290 E-mail: [email protected] M.Sc. Marcus Mergenthaler, Institute for Agricultural Economics and Social Sciences in the Tropics and Subtropics (490), University of Hohenheim, 70593 Stuttgart, Germany; Phone: +49 (711) 459 2602, Fax: +49 (711) 459 3762 E-mail: [email protected] 106 CONTACT ADDRESSES

Dr. Klaus Meyn, ADT Projekt GmbH, Adenauerallee 174, 53113 Bonn, Germany; Phone: +49 (228) 914 4730, Fax: +49 (228) 914 473 1593; E-mail: [email protected] Dr. Helmut Momm, Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart, Germany; Phone: +49 (711) 457 9891, Fax: +49 (711) 457 9891 E-mail: [email protected] M.Sc. Katinka Musavaya, Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart, Germany; Phone: +49 (711) 459 3006, Fax: +49 (711) 459 3290 E-mail: [email protected] Dipl. Ing. agr. bio. Regina Roessler, Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart, Germany, Phone: +49 (711) 459 4247, Fax: +49 (711) 459 3290 E-mail: [email protected] Dipl. Ing. agr. Tobias Rummel, Dorfstr. 38, 09326 Geringswalde, OT Neuwallwitz, Germany; Phone: +49 (3738) 271 722 Dr. Cornelia Schäfer, Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart, Germany; Phone: +49 (711) 459 3006, Fax: +49 (711) 459 3290 E-mail: [email protected] Dr. Angelika Stemmer, Casilla 1879, Cochabamba, Bolivia; Phone: +59 (1) 421 8725 E-mail: [email protected] Prof. Dr. Anne Valle Zárate, Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart, Germany; Phone: +49 (711) 459 3170, Fax: +49 (711) 459 3290 E-mail: [email protected]

7.2 Advisory panel Prof. Dr. John Gibson, Director, The Institute for Genetics and Bioinformatics, Hawkins Homestead, University of New England, Armidale, NSW 2351, Australia; Phone: +61 (02) 6773 2930, Fax: +61 (02) 6773 3275, Mobile: +61 (0437) 039951 E-mail: [email protected] Dr. Irene Hoffmann, Chief, Animal Production Service, Animal Production and Health Division, FAO, Viale delle Terme di Caracalla, 00153 Roma, Italy; Phone: +39 (06) 570 52796, Fax: +39 (06) 570 55749 or -53927, Mobile: 348 (870) 5309 E-mail: [email protected] CONTACT ADDRESSES 107

Annette von Lossau, Project officer, GTZ-Project "People and Biodiversity", P.O. Box 5180, 65726 Eschborn, Germany; Phone: +49 (6196) 79 1473 E-mail: [email protected] Dr. Elzbieta Martyniuk, Department of Animal Genetics and Breeding, Warsaw Agricultural University, ul. Ciszewskiego 8, 02-786 Warszawa, Poland Phone/Fax: +48 (22) 853 09 31 E-mail: [email protected] Evelyn Mathias, League for Pastoral Peoples and Endogenous Livestock Development, Weizenfeld 4, 51467 Bergisch Gladbach, Germany; Phone: +49 (2202) 932921, Fax: +49 (2202) 932922 E-mail: [email protected] Dr. Siboniso Moyo, Dept. of Livestock Production, Bevan Building, 18 Borrowdale Road, Harare, Zimbabwe Phone: +263 (04) 731126 or 702584 E-mail: [email protected] Dr. Ferdinand Schmitt, Managing Director, ADT Projekt GmbH, Adenauerallee 174, 53113 Bonn, Germany; Phone: +49 (228) 91447 32, Fax: + 49 (228) 91447 31, Mobile:+ 49 (171) 286 1034 E-mail: [email protected]

7.3 Resource persons Abdi Ahmed, Pastoral Forum Ethiopia (PFE), Panos, POB 1570 code 1110 or 1152, Addis Ababa, Ethiopia; Phone: +251 (11) 666363/59, Fax: +251 (11) 666361 E-mail: [email protected] Dr. Jock Allison, Abacus Biotech, PO Box 5585, Dunedin, New Zealand; Phone: +64 (3) 4776375, Fax: +64 (3) 4776376, Mobile 64 (21) 363337 E-mail: [email protected] Arbeitsgemeinschaft Österreichischer Fleckviehzüchter (AGÖF), Pater Werner Deibl Straße 4, 3910 Zwettl, Austria, www.fleckvieh.at Phone: +43 (2822) 53531-10, Fax: +43 (2822) 53531-15 E-mail: [email protected] Awassi Rt., 4164 BAKONSZEG, Hunyadi út 83., Hungary; Phone: (54) 513-000/001/002, Fax: (54) 513-003, www.awassi.hu E-mail: [email protected] Workneh Ayalew, ILRI-BMZ Project Manager, Animal Genetic Resources Group, International Livestock Research Institute (ILRI), POB 5689, Addis Ababa, Ethiopia; Phone: +251 (11) 6463215, Fax: +251 (11) 6461252/ 6464645 E-mail: [email protected] M. Bass Werner, Director of DRIPAD Tarija, Avenida Defensores del Chaco s/n, zona aeropuerto, Tarija, Bolivia; Phone: +591 (46) 645837 E-mail: [email protected] 108 CONTACT ADDRESSES

Dr. Maurice Bichard, The Farmhouse, Fyfield Wick, Oxon, OX13 5NB, Abingdon, United Kingdom; E-mail: [email protected] Prof. Emeritus Rigoberto Calle Escobar, Universidad Nacional Agraria La Molina, Telf. 3495747 anexo 164., Francisco Del Castillo 573 San Antonio, Miraflores, Lima -Perú; Phone: 4473115, Fax: 2424364, Mobile: 99857639 E-mail: [email protected] F. Cautin, Administrator of SONU (Sociedad Nueva), Calle Mama Ocllu No. S-0905, Casilla 4231, Cochabamba, Bolivia; Phone: +591 44250562 E-mail: [email protected] Yosef Carrasso, Sheep and Goat Division, Extension Service, Ministry of Agriculture and Rural Development, P.O. Box 28, Bet Dagan 50250, Israel; Phone: 97239485306, Fax: 97239485614, Mobile: 97252993656 E-mail: [email protected] E. Chumacero, Faculty of , University Tomas Frias, Avenida del Maestro s/n, Potosi, Bolivia; Phone: +591 46225414 Morné De la Ray, EMBRYO PLUS, POB 2644, BRITS, 0250, North West, South Africa Phone: +27 (12) 250 2359, Fax: +27 (12) 250 2299, http://www.embryoplus.com E-mail: [email protected] Dr. Nguyen Van Dong, Pig research centre, National Institute of Animal Husbandry, Chem, Tu Liem, Hanoi, Vietnam; Phone: +84 (4) 938 9774, Mobile: +84 (9) 13001340 Ebru Emsen, Department of Animal Science, Ataturk University, Turkey E-mail: [email protected] R. Ergueta, Formerly technician of CEDEAGRO, Mizque, Cochabamba, currently Director of DRIPAD Cochabamba, Avenida Aroma No. 327, Cochabamba, Bolivia; Phone: +591 44251565 E-mail: [email protected] Irina Florescu, Counsellor, Division for European Integration, Ministry of Agriculture, Romania; E-mail: [email protected] C. Foyanini, Owner of Residencial Bolivar and goat farms in Pailon and San Javier, Santa Cruz, Bolivia; Phone: +591 3325989 E-mail: [email protected] A. Fraga, Diretor Presidente, Associacao Brasileira de Criadores das Racas Simental e Simbrasil, Rua Mario Romanelli, 23 - Bairro Gilberto Machado / Cachoeiro de Itapemirim- ES - ECP: 29303-260, Brasil; Phone: +28 (3521) 5666, Fax +28 (3521) 0570 E-mail: [email protected] CONTACT ADDRESSES 109

Tezera Getahun, Executive Director, Pastoral Forum Ethiopia (PFE), Panos, POB 1570 code 1110 or 1152, Addis Abeba, Ethiopia; Phone: +251 (11) 666363/59, Fax: +251 (11) 666361 E-mail: [email protected] and [email protected] Phillip Grand, Cowra Cheese, 1078 Mid Western Highway, Cowra, NSW 2794 Australia; Phone: +61 0417479716 R. Hinojosa, National Director of Heifer International Bolivia, Avenida Moscu s/n entre 5 y 6 anillo, Santa Cruz, Bolivia; Phone: +591 33557235 E-mail: [email protected] O. Hinojosa, Technician of Aldeas SOS, Cochabamba, km 1 a Tiquipaya, zona Linde s/n, Cochabamba, Bolivia; Phone: +591 44421935 E-mail: [email protected] Dr. Sandor Kukovics, Research Institute for Animal Prod. and Nutrition, Geszteneyés u.1. Herceghalom, H-2053 Hungary; Phone: +3623319133 E-mail: [email protected] Dr. Nurlan Malmakov, Research Institute of Sheep Breeding, Research Centre for Animal Production and Veterinary, Mynbaevo village, Djambul District, Almaty region, 483174, Kazakhstan; Phone: +73272214235 E-mail: [email protected] Geoff Maynard, Mount Eugene Belmont Reds & 5 Star Senepols, "Mount Eugene", POB1, Jambin 4702, Queensland, Australia; Phone: +61 (7) 4996 5240, Fax: +61 (7) 4996 5316 E-mail: [email protected] Ian McDougall, 48 Bath Road, Stroud GL5 3JL, United Kingdom; Phone:+44 7855262308 E-mail: [email protected] B. Mendoza, Lecturer at the National University of Loja, Ciudadela Universitaria, La Argelia Casilla 795, Loja, Ecuador; Phone: +593 7574054 E-mail: [email protected] Tafesse Mesfin, FARM-Africa Ethiopia, POB 5746, Addis Ababa, Ethiopia; Phone: +251 (11) 155 26 84, Fax +251 (11) 155 21 43 E-mail: [email protected] Dr. Charlotte Milne, Dorper Sheep Breeders’ Society of South Africa, PO Box 26, 42 Van Reenen Street, Middelburg ECP, 5900, South Africa; Phone: +27 (49) 842 2241, Fax: +27 (49) 842 3589 E-mail: [email protected] 110 CONTACT ADDRESSES

Joaquín P. Mueller, INTA, Estación Experimental Agropecuaria Bariloche. CC 277 (8400) Bariloche, Río Negro, Argentina; E-mail: [email protected] Andrew Mushita, Commutech, POB 7232, Harare, Zimbabwe; Phone: +263 (4) 589 256, Fax: +263 (4) 589 390 E-mail: [email protected] Chanda Nimbkar, Nimbkar Agricultural Research Institut, Animal Husbandry Division, P.O. Box 23, Phaltan 415523, Maharashtra, India; E-mail: [email protected] Ch. Papachristoforou and C. Christofides, Ministry of Agriculture, 1412 Nicosia, Cyprus; Phone: +35722408639, Fax: +35722408656 E-mail: [email protected] and [email protected] Maria Norma Ribeiro, Departamento de Zootecnia, Universidade Federal Rural de Pernambuco, Av. D. Manoel de Medeiros, SN, Dois Irmaos, CEP: 51171-900 Recife, Brazil; E-mail: [email protected] Daniel Roldao, Sociedade Agricola da Herdade do Matinho, Lda. P.O. Box 22. Quinta da Moutosa,7320 Castelo de Vide, Portugal; www.herdadematinho.pt Phone: +351 245901145/ 245993225 / 245202202, Fax: +351 245901145, E-mail.: [email protected], [email protected] Dr. Hermann Schulte-Coerne, Referat 322 - Tierzucht und Tierhaltung- Bundesministerium für Verbraucherschutz, Ernährung und Landwirtschaft, Rochusstr. 1, 53113 Bonn, Germany; Phone: +49 (1888) 529 3484, Fax: +49 (1888) 10 529 3484 E-mail: [email protected] Dr. David Steane, 99 Moo 7 Baan Rong Dua, Tha Kwang, Saraphi, Chiang Mai 50140, Thailand; Phone/Fax: +66 (53) 429- 918 E-mail: [email protected] Makros Tibbo, International Livestock Research Institute ILRI Animal Genetic Resources, PO Box 5689, Addis Ababa, Ethiopia; E-mail: [email protected] Dr. Le Thi Thuy, Animal Genetic Molecular Laboratory, National Institute of Animal Husbandry, Chem, Tu Liem, Hanoi, Vietnam; Phone: +84 (4) 838 9165, Fax: +84 (4) 8389775 E-mail: [email protected] J.L. Vaca, Lecturer at the Veterinary Faculty of University Gabriel Rene Moreno, Avenida Centenario, Campus Universitario, Casilla 702, Santa Cruz, Bolivia; Phone: +591 3537676 E-mail: [email protected] Dr. John E Vercoe AM FTSE, 7 Ryan Street, Zilzie Q 4710, Australia; Phone/Fax: +61 (7) 49387486 E-mail: [email protected] CONTACT ADDRESSES 111

Jacob Wanyama, ITDG-EA (Intermediate Technology Development Group East Africa), POB 39493, Postal Code: 00623, Nairobi, Kenya; Phone: +254 (2) 2713540, Fax: +254 (2) 2710083 E-mail: [email protected] Kerstin Zander, Center for Development Research (ZEF), Walter-Flex-Straße 3, 53113 Bonn, Germany; Phone: +49 (228) 731852 E-mail: [email protected] 112 ACKNOWLEDGEMENTS

8 ACKNOWLEDGEMENTS

This report was written by a team of authors and editors with whom it was a pleasure to work. Special thanks are extended to the senior experts in our team, Prof. em. Dr. Christian Gall and Dr. Helmut Momm, who were available at all times to write, review or assist in all matters. Special thanks are also extended to the advisory panel comprising of John Gibson (University of New England, Australia), Irene Hoffmann (FAO), Annette von Lossau (GTZ), Elzbieta Martyniuk (National Coordinator of Poland), Evelyn Mathias (League for Pastoral Peoples), Siboniso Moyo (National Coordinator of Zimbabwe) and Ferdinand Schmitt (ADT Projekt GmbH) for accompanying the study with vigour and expertise. Thank you for your presence at the meetings, reading of drafts and valuable inputs throughout the study. We are very grateful to individual experts who read parts of the study and gave helpful comments on the interpretation of data accumulated in the study. We would especially like to mention Maurice Bichard, Elisha Gootwine, Klaus Meyn, David Steane and John Vercoe. Furthermore, we highly appreciate and acknowledge all contributions from resource persons listed in the previous chapter for their patience to attend to our questions and provide valuable information. 113

GENE FLOW IN ANIMAL GENETIC RESOURCES. A STUDY ON STATUS, IMPACT AND TRENDS

Editors: Anne Valle Zárate, Katinka Musavaya and Cornelia Schäfer

Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, Germany

ANNEX 9.1

114

ANNEX 115

9 ANNEX

Annex 9.1 Background information 116 TABLE OF CONTENTS

TABLE OF CONTENTS

Abbreviations 118

Background information 119 A 1: Advisory panel 119 A 2: Development of the Merino breed in Europe 119 A 3: Current gene flow in sheep indicated by selected export and import records and records on introduction of foreign genetic material 121 A 4: Composite sheep breeds, number of foundation breeds and the year of origin or year recognised by country of origin 122 A 5: ICAR Report of the working group on milk recording scheme 128 A 6: Changes in sheep breed composition in the USA 132 A 7: Country Report excerpts for sheep 133 A 8: Country Report excerpts for goats 136 A 9: Possible introduction routes of cattle from West to South, Southeast, East and Northeast Asia, with centres of domestication 139 A 10: Possible introduction routes of cattle into North, West and northeastern Africa 139 A 11: Possible introduction routes of cattle into the Indian subcontinent and Southeast Asia and probable centres of domestication for Bos (bibos) spp. 140 A 12: Cattle migration after the 16th century 141 A 13: Total Simmental population per country** in the early 1990s 141 A 14: German Simmental export to Turkey 142 A 15: EU heifer export by origin and year 142 A 16: Annual EU heifer export by destination 1993-2003 143 A 17: Annual EU heifer import and export 1998-2002 144 A 18: EU heifer import by destination and year 144 A 19: Annual EU heifer import by origin 1993-2003 145 A 20: German Simmental export by year 145 A 21: Introduction of the Simmental breed to different environments 146 A 22: Crosses of Simmental with local cattle breeds 147 A 23: German Simmental export to Balkan countries 1998-2002 147 A 24: Semen export of industrialised countries by origin in 1991 148 A 25: EU semen export by destination in 2003 (%) 148 A 26: EU semen export by destination in 2003 (number of doses) 149 TABLE OF CONTENTS 117

A 27: Semen import of developed countries by origin in 1991 151 A 28: EU semen import by origin in 2003 152 A 29: Semen imports and exports by breed group and region in 1991 153 A 30: Semen imports to developing countries in 1991 by breed, breed group and destination 153 A 31: Annual semen imports and exports of EU in 2002/2003 154 A 32: German semen import (doses) by year and origin 154 A 33: US Brahman export by destination and year 155 A 34: Brazilian Simmental embryo import by origin 1986-1993 155 A 35: Brazilian Simmental semen import by origin 1972-1993 156 A 36: Brazilian semen doses sold in 2002 by breed 156 A 37: Israeli Holstein export by country and time period 157 A 38: AI coverage by breed and country groups for developed countries in 1991 157 A 39: AI coverage by breed and country groups for developing countries in 1991 158 A 40: Use of AI by breed group and region for developed countries in 1991 158 A 41: Use of AI by breed group and region for developing countries in 1991 159 A 42: Share of bull breeds used in Botswana from 1987 to 1995 159 A 43: Australian beef cattle registrations by breed 160 A 44: Number of breeders by breeds in South Africa in July 2003 161 A 45: Ratio of utilisation of the Simmental breed in different countries for milk and beef 161 A 46: Examples of absorption of local cattle by imported breeds 162 A 47: Holstein’s and Sahiwal’s contribution to composite breeds 162 A 48: Performance of the Holstein breed in different environments 163 A 49: Country Report excerpts for cattle 164 A 50: Annual EU breeding pig import and export 1992-2003 172 A 51: Imported pig breeds during the 1970s into different countries 173 A 52: State of the pig genetic resource utilisation in Uruguay 174 A 53: Country Report excerpts for pigs 175 A 54: Information from the World Watch List of Domestic Animal Diversity 182

References background information 183 118 ABBREVIATIONS

ABBREVIATIONS

ADT Projekt Specialised consulting enterprise that focuses on planning and implementing international agricultural projects (Germany) AI Artificial insemination e.V. Eingetragener Verein (Incorporated society) EU European Union EU-15 Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Portugal, Spain, Sweden, The Netherlands, United Kingdom FAO Food and Agriculture Organisation of the United Nations FARM-Africa Food and Agricultural Research Management Organisation (UK) GmbH Gesellschaft mit beschränkter Haftung (limited liability company) GTZ Gesellschaft für Technische Zusammenarbeit i.e. that is ICAR International Committee for Animal Recording IFAD International Fund for Agricultural Development n.a. not applicable NGO Non-Governmental Organisation No. number PIC Pig Improvement Company Limited (UK) UK United Kingdom US United States USA United States of America USSR Union of Socialist Soviet Republics

BACKGROUND INFORMATION 119

BACKGROUND INFORMATION

A 1: Advisory panel

Irene Hoffmann FAO, Chief, Animal Production [email protected] Service Annette von Lossau GTZ, Sectoral Project "People and [email protected] Biodiversity in Rural Areas" Elzbieta Martyniuk FAO, National Coordinator of [email protected] Poland Evelyn Mathias NGO, League for Pastoral Peoples [email protected] Siboniso Moyo FAO, National Coordinator of [email protected] Zimbabwe Ferdinand Schmitt ADT Projekt GmbH, Managing [email protected] Director

A 2: Development of the Merino breed in Europe

In Germany, the year 1765 marks the year of the German Merino breed when Saxony first sought the right to buy Merino breeding stock from Spain. Then, 92 rams and 128 Merino ewes were brought to Saxony where they were crossbred with local indigenous breeds. Further imports of Merino breeding stock from Spain established the breed in Saxony. In Lohmen, from 1883 to 1923, a purebred flock was kept and the combination of wool quality with wool quantity was realised. At that time, the Merino population in Saxony counted 2 million animals and their wool was highly demanded. At the same time, Merino populations developed in Prussia and South Germany by introducing pure Merinos from Spain and crossbreeding with local indigenous breeds. Later, imports from Saxony and Austria followed (Haring, 1984). In Austria, the Merino breed was introduced by Maria Theresa with 400 ewes brought from Spain before 1784. Further imports followed and established the Austrian Merino breed, which, from the Wool convent in Leipzig in 1823 onwards, was called “Negretti”. In France, Spanish Merinos arrived as a gift from Spain in 1786 to Rambouillet and their rams were crossbred with local landrace ewes to improve wool quality and quantity. The Rambouillet breed was founded, which can still be found purebred in a flock of 130 ewes (Strittmatter, 2003). From 1850 until today, all exports of breeding stock from this flock have been recorded and therefore gene flow can be documented in an exemplary way (Haring, 1984). Exports to many countries took place. Firstly, breeding stock was exported to North America in 1850, where the largest Merino-Rambouillet population outside France still exists (Haring, 1984). America was followed by Australia in 1863, Argentina in 1864, Germany and Russia in 1867, Uruguay in 1869, South Africa in 1870, New Zealand in 1873, England in 1897 and Israel in 1959 (Haring, 1984). The Merino d’Arles developed in parallel in Perpignan and Arles. Detailed on the development of the Merion breed in eastern European countries was only 120 BACKGROUND INFORMATION available for Hungary. Hungary was for many years a major route in the Great Migration (Fesüs et al., 2002). Since the movements were from East to West, it may be assumed that most of the sheep found in Hungary were originally eastern. Merino sheep breeding began in 1774 in Hungary when Maria Theresa bought 300 Merinos from Spain (Fesüs et al., 2002). The geography of the Carpathian Basin was favourable to the spread of the Merino breed. The local Merino breeding program commenced in 1786, when the Emperor Joseph II removed the prohibition on private individuals bringing Spanish sheep into the country and made their import duty free. The heyday of wool-oriented Merino sheep breeding lasted a full 100 years, beginning in the middle of the 18th century and lasting to the middle of the 19th century. In the milder regions of South Africa, the South African Merino gained importance when South Africa became the first country outside Europe to own Merinos in 1789. From then onwards, the Merino breed spread from the well established sheep farming areas in the Western and South-western Cape eastwards and with the Great Trek in 1834 northwards. Within a few years the Merino breed, due to the suitable environment, had spread to all parts of South Africa. Then two Spanish Merino rams and four Spanish Merino ewes were donated by the Dutch Government to the military commander, Col Jacob Gordon, at the Cape on an experimental basis. Initially, the King of Spain had sent a number of sheep from his famous Escoriale Merino Stud as a gift to the house of Orange. But since the Merino could not adjust to the high rainfall in the Netherlands, they were sent to the Cape. Here, Col Gordon realised the potential of the breed and kept it pure. By 1830, wool sheep farming in the Western and South-western Cape was already well established. The next expansion went eastwards and the 1820 settlers played an important role in extending and developing the Merino flocks in South Africa. In 1834 the Great Trek started, which took their sheep flocks northwards with them. Within a few years the Merino had spread to all parts of the country. From 1891 considerable numbers of different types of Merino were imported from other continents to improve the flocks. The Australian Merino, the Wanganella and Peppin type, were found to suit best and large numbers were imported to form the basis of the typical South African Merino (Merino Breeders’ Society of South Africa, 2004). In the Highlands of Kenya, the Corriedale became an important breed which developed from the South African Merino sheep by crossing them with Lincoln and Leicester. This breed spread from South Africa to Chile-Patagonia, USA, Canada and Kenya (Hammond et al., 1961).

BACKGROUND INFORMATION 121

A 3: Current gene flow in sheep indicated by selected export and import records and records on introduction of foreign genetic material

Breed Country of origin Country of destination Kent Not recorded Poland Leine Not recorded Poland East Friesian Not recorded Poland Finn Not recorded Poland Lincoln Not recorded Poland Booroola Merino Not recorded Poland Texel Not recorded Poland Ile de France Not recorded Poland Blackheaded Mutton Sheep Not recorded Poland Suffolk Not recorded Poland Berrichon du Cher Not recorded Poland Dorset Not recorded Poland Merino Australia Russia Rambouillet USA Russia Cameroon West Africa Germany Dorper South Africa Germany Chios Greece Near East (countries unrecorded) Dorper South Africa USA East Frisian Not recorded USA Lacaune Not recorded USA Several unrecorded breeds Canada, France USA Barbados Blackbelly Central America USA Texel Not recorded USA St. Croix USA Canada, Mexico Hamshire Down Not recorded Uruguay South Down Not recorded Uruguay Suffolk Not recorded Uruguay Ile de France Not recorded Uruguay Texel Not recorded Uruguay Poll Dorset Not recorded Uruguay Dohne Merino Australia Uruguay Merino Australia Uruguay New Zealand Uruguay Corriedale Australia, New Zealand Uruguay Ideal Australia Uruguay Romney Marsh New Zealand Uruguay Merino Australia, New Zealand Argentina Dorper South Africa Mexico, Brazil 122 BACKGROUND INFORMATION

Breed Country of origin Country of destination Barbados Blackbelly Not recorded Ecuador Pelibuey Not recorded Ecuador Dorper South Africa Australia 25 breeds including Polled Russia, France, UK, Germany, China Dorset, Suffolk, Texel, Netherlands, Denmark, USA, Charolais Canada, Australia, New Zealand Djallonke Germany Malaysia Barbados Blackbelly USA Indonesia St. Croix USA Indonesia Improved Awassi Israel Details given in case study D’man Morocco Iraq Breeds not recorded EU-15 countries EU-15 countries, Balkan countries, Switzerland, Turkey, Morocco, Algeria, Venezuela, Brazil, China, Japan Breeds not recorded Switzerland, Poland, Chile, EU-15 countries Argentina, Uruguay Source: own elaboration, summary of chapter 3 on current status of gene flow of sheep

A 4: Composite sheep breeds, number of foundation breeds and the year of origin or year recognised by country of origin Country Composite No. of Country Composite No. of breed foundation breed foundation population breedsa population breedsa Algeria Tadmit 2 (1925) Dormer 2 Argentina Argentine 4 (1979) Elliotdale 4 (1963) Cormo Fonthill Merino 2 (1954) Corino 2 (1970) Glenara 3 Pampinta 2 (1980) Improver Armenia Armenian 3 Gromark 2 (1979)b Semicoarsewool Hyfer 3 (1978) Kyasma 2 Improved 2 Martunin 3 Australia BLM 2 (1955) Meridale 24 Bond 3 (1909) Poll Dorset 2 (1954)b Booroola 2 Polwarth 2 (1880) Leicester Romshire 2 Borino 2 Waridale 3 (1970) Bumfdale 5 (1979) White Suffolk 3 (1977) Bundoran 2 (1971) Zenith 2 (1947) Comeback Austria Austrian 2+ Bungaree 2 Negretti Merino Carinthian 3 (1988)b Comeback 2 (1976)b Azerbaijan Azerbaijan 3 (1947) Coolalee 6 (1968) Mountain Cormo 2 (1960) Merino Daldale 3 (1970) Brazil Brazilian Somali 2 (1939) BACKGROUND INFORMATION 123

Country Composite No. of Country Composite No. of breed foundation breed foundation population breedsa population breedsa Brazilian 2 Ningxia Black 2 Woolless North-East 3 Rabo Largo 2 China Finewool Santa Ines 2 (1940) North-East 2 Burkina Mossi 2 China Faso Semifinewool Bulgaria Bulgarian Dairy 2 (1970) 4 Danube 4 (1950) Semifinewool Finewool Shanxi Finewool 2 (1920) Karnobat 2 (1950) Sichuan 4 (1970) Finewool Semifinewool Mountain Tsigai 3 (1950) Tibegolian 2 North Bulgarian 6 4 (1935) Semifinewool Finewool North-East 4 (1950) Yunan 2 (1970) Bulgarian Semifinewool Finewool Colombia Manchada 2 (1976) Petrokhan Tsigai 3 Paramuna Pleven 2 Croatia Dubrovnik 2 (late Blackhead 18th) Razlog 3 (1955) Island Pramenka 2 South Bulgarian 4 Pag Island 2 (19th) Semifinewool Czeck Improved 4 Thrace Finewool 6 (1943) Republic Valachian White 3 (1916) Sharka 5 Klementina Denmark Danish 2 Zlatusha 3 (1965) Finewool Cameroon Maroua 2 Danish Landrace 4 (1900) Canada Canadian Arcott 12 (1988)b Såne 3 (1991) Canadian 3 (1919) Egypt Ossimi-Finn 2 (1980) Corriedale Rahmani-Finn 2 (1980) DLS 3 (1989) Estonia Estonian 2 (1940) Outaouais 9 (1988)b Darkheaded Arcott Estonian 2 Newfoundland 7 (19th) Whiteheaded Rideau Arcott 9 (1988)b France Avranchin 4 (1928)b Romnelet 2 (1935) Berrichon du 6 (1936)b China Aohan Finewool 3 (1970) Cher Chinese Karakul 3 (1960) Blanc du Massif 2 (1965) Chinese Merino 3 Central Erduos 3 Bleu du Maine 3 (1938)b Gadasu 2 (1970) Boulonnais 3 (1963) Alpine 6 (mid Catalan 4 Finewool 20th) Central 2 Guizhou 5 Pyrenean Mutton-Wool Charmoise 5 (1896)b Inner Mongolian 2 Charollais 2 (1963)b Finewool Cotentin 2 (1925)b Jia Shike 2 (1970) French Alpine 2 (1952)b Lanzhou Large- 2 (1862) FSL 3 (1967) tail Ile de France 2 (1922)b Linchuan 3 (Dishley 124 BACKGROUND INFORMATION

Country Composite No. of Country Composite No. of breed foundation breed foundation population breedsa population breedsa Merino) Bharat Merino 5 (1980) INRA 401 2 (1980) Hissardale 2 (19th) Landais 2 Kashmir Merino 7 (1947) Lourdais 2 (1975)b Nilgiri 4 (18th) Rayole 2 (1980) Raymond 5 (1973) Roussillon Red 3 Merino Roussin de la 4 (1983) Sandarsamand 2 (1935) Hague Sandyno 2 (1973) Trun 2 (1960) UAS 3 Georgia Georgian Fat- 3 (19th) Indonesia Priangan 3 (19th) tailed Finewool Ireland Belclare 4 (1985)b Georgian 4 (1931) Improver Semifinewool Fingalway 2 (1970) Fat-tailed Finn-Dorset 2 Germany Bentheimer 2 (1934)b High Fertility 6+ (1965) German 4 (1920)b Improved 2 Blackheaded Galaway Mutton Israel Assaf 2 (1955) German Karakul 2 Israeli Improved 2 (1943)b German 3 (1938)b Awassi Mountain Italy Apennine 3 (1981)b German 3 (1885)b Campanian 2 (1971)b Whiteheaded Barbary Mutton 2 (1942)b b Leine 5 (1906) 3 Merino 3 (1971) 2 (1974)b Longwool 2 b Merinolandschaf 2 (1922) 4 (1942)b Ghana Nungua 2 Mascherina 2 Blackhead Segezia Triple 3 Greece Argos 2 Cross Evdilon 2 Sicilian Barbary 2 (1942) Frisarta 4 (1946) 5 (1942) Moraitiko 2 Tyrol Mountain 3 Mytilene 2+ Kazakhstan Aktyubinsk 2 Rhodes 2 Semicoarsewool Greenland Greenland 2 Chuisk 4 Hungary Csenger Merino 2 (1982) Semifinewool Hungarian 7 Degeres Mutton- 3 (1931) Merino Wool Hungarian 2 (1992) Edilbaev 2 Prolific Merino Kargalin Fat- 4 Kazanluk 5 (1964) rumped Semifinewool Kazakh Arkhar- 2 (1934) J-AKI-1 2 (1980) Merino J-AKI-2 3 (1980) Kazakh 3 Prolific Babolna 3 (1970) Corriedale Tetra 3 (1960) Kazakh 3 (1946) Iceland Kleifa 3 Finewool India Avikalin 2 (1970) Kazakh 5 (1945) Avivastra 2 Semifinewool Baghdale 3 North Kazakh 2 (1976) BACKGROUND INFORMATION 125

Country Composite No. of Country Composite No. of breed foundation breed foundation population breedsa population breedsa Merino Halfbred South Kazakh 6 (1966) New Zealand 2 (1974) Merino Wiltshire West 5 (1952) Perendale 2 (1961)b Kazakhstan Romney- 2 Mutton-Wool Corriedale Kyrgyzstan Kirgiz Fat- 3 Skye Farm 3 (late rumped Romney 1960) Kirgiz Finewool 4 (1956) South Dorset 2 Tyan Shan 5 (1966) Down Latvia Latvian 4 (1937) South 2 (1970) Darkheaded Hampshire Libya Barbary 2 South Suffolk 2 (1938) Halfbred Nigeria Permer 2 Ghimi 2 Yankasa 2 Lithuania Lithuanian 3 (1923) Norway Dala 3 (1926)b Blackheaded Norwegian Fur 2 (1968)b Mexico Chiapas 5 sheep Tarset 2 Rygja 3 (1926)b Mongolia Aohan Finewool 3 (1970) Steigar 2 (1954) East Mongolian 4 (1962) Pakistan Baghdale 3 Semifinewool Pak Awassi 2 Hangay 2 Pak Karakul 2 (1965) Orhon 4 (1943) Peru Asblack 3 (1990) Sumber 2 (1950) Junin 5 (1940) Torguud 2 (1962) Philippines Laguna 2 Yoroo 4 (1981) Poland Bialystok 2 (1963) Morocco South 2 Bochnia 2 Morroccan Damline 66 3 (1980) Timahdit 3 Damline 77 3 (1980) Zaian 2 Friserra 2 (1960) Zoulay 2 Jędrzychowice 2 (1954) The Dutch Black 4 (1979)b Merino Netherlands Blaze Kamieniec 3 (1954) Flevoland 2 (1975) Koszalin 4 North Holland 2 (1970) Lublin 3 (1950) Rijnlam A 2 Olkusz 2 Rijnlam B 3 Polish 4 (1976) Schoonebeker 2 (1990)b Blackheaded Swifter 2 (1967) Mutton Texel 3 (1909)b Polish 2 (1962) New Borderdale 2 (1930) Corriedale Zealand Polish 5 Border-Merino 2 Longwool Border-Romney 2 Polish Lowland 9 Carpetmaster 3 Polish Merino 2 (1952) Cheviot- 2 Polish Mountain 3 (1946) Corriedale Polish 3 Coopworth 2 (1968)b Strongwooled Corriedale 2 (1910)b Merino Elliotdate 4 (1960) Polish 6 (1976) New Zealand 4 Whiteheaded 126 BACKGROUND INFORMATION

Country Composite No. of Country Composite No. of breed foundation breed foundation population breedsa population breedsa Pomeranian 4 (1984) Russian 2 (1978) Prolific 09 3 (1976) Longwool Prolific meat 08 5 (1976) Russian 2 Prolific meat 10 8 (1976) Mountain Prolific wool 04 2 (1976) Merino Sireline 5 (1970) Salsk 2 (1932) Silesian 2 (1932) Siberian Merino 5 (20th) Wielkopolska 2 (1977)b Siberian type 3 Żelazna 3 (1955) Soviet Mutton- Portugal Bordaleiro 2 Wool Fonte Bôa 2 (1902) Soviet Merino 6 (1938) Merino Soviet Mutton- 5 (1950) Friserra 2 (1962) Wool Portuguese 4 (1929) Stavropol 4 (1950) Merino Transbaikal 6 (1927) Romania Bîrsa 3 (1953) Finewool Danube Merino 2 Volgograd 6 (1978) Palas Merino 6 (1926)b Vyatka 2 (1936) Ruşeţu 1 3 Serbia Vojvodina 2 (19th) Spancǎ 2 Merino Stogoşǎ 2 Senegal Warale 2 (1975) Russia Altai 4 (1940) Slovakia Improved 4 Altai Mountain 3 (1945) Valachian Angara Merino 3 (1974) Slovakian 3 Merino Caucasian 3 (1921) b Chita 2 Solčava 3 (1983) Dagestan 2 (1926) South Afrino 3 (1969) Mountain Africa Gorki 2 (1950) Bezuidenhour 3 (1918) Grozny 2 (1929) Africander Kalinin 2 (1935) Dohne Merino 2 (1940) Karachai 3 Dormer 2 (1941) Dorper 2 (1950)b Mountain b Mutton-Wool South African 5 (1906) Merino Krasnoyarsk 5 (1963) b Finewool Van Rooy 3 (1948) b Kuchugury 2 (1970) White Dorper 2 (1960) Kuibyshev 2 (1938) White Wooled 3 (1942) Kulunda 3 Mountain Liski 2 (1936) Spain Basco-Béarnais 2 (1960) Manych type of 2 Mestizo 2 Stavropol Entrefino-fino North Caucasus 3 (1944) Ripollesa 2 Semifinewool Salz 2 (1970) North Caucasus 3 (1944) Talaverana 3 (1960) Mutton-Wool Sudan Ingessana 2 Omsk 2 Meidob 2 Semifinewool Toposa 2 b Oparino 2 Switzerland Engadine Red 2 (1985) b Ostrogozhsk 2 (1963) Swiss 2 (19th) Pechora 2 (1937) Brownheaded Swiss Charollais 2 (1991)b BACKGROUND INFORMATION 127

Country Composite No. of Country Composite No. of breed foundation breed foundation population breedsa population breedsa Swiss White 2 (1936) Improver Alpine 9 (1969)b Togo Vogan 2 Castlemilk 3 (1974)b Tajikistan Darvaz 3 (1948) Moorit Mountain Chevaldshay 2 Mutton-Wool Clun Forest 3 (1865) Pamir Finewool 2 Colbred 4 (1962)b Tajik 3 (1963) Cotswold 2 (1862) Tunisia Sidi Tabet cross 2 Crickleg Barrow 2 (1970) Tadmit 2 (1925) 2 (1930) Thibar 2 (1945)b Dartmoor 2 (1909) Tunisian milk 2 Derbyshire 2 (1892)b sheep Gritstone Turkey Acipayam 3 Devon, 2 (1977) Çandir 3 Cronwall Central 2 (1952) Longwool Anatolian Devon 2 (1923)b Merino Closewool Kamakuyruk 2 2 (1906)b Kivircik 2 Dorset Horn 2 (1862) Karacabey- 2 (1928) English 2 (1981)b Merino Halfbred Menemen 3 Greyface 2 Menemen 2 Oldenbred Kivircik Hampshire 3 (1889)b Ramliç 2 (1969) Down Tahirova 2 (1964) Kent Halfbred 2 (1988)b Türkgeldi 3 (South England) Ukraine Azov Tsigai 2 Lincoln 2 (1892)b Askanian 4 Longwool Blackheaded Llanwenog 2 (1963)b Askanian 2 Lleyn 3 (1970)b Corriedale Masham 2 Askanian 3 Meatlink 5 (1963) Crossbred 4 (1978)b Large Karakul 2 (1932) North of 2 (1980)b Multifoetal 2 (1935) England Mule Karakul Oldenburg 2 North Ukrainian 4 3 (1951) Semifinewool Pettadale 2 (1959) Transcapathian 4 Romney 2 Finewool Halfbred Ukrainian 2 (1950) Scotch Halfbred 2 Mountain Scotch Mule 2 (1986)b United ABRO Damline 4 (1967) Scottich 2 Kingdom Greyface Black Leicester 2 (1986)b Scottish 2 Longwool Masham Boreray 2 (1930) Shetland- 2 British 5+ (1970) Cheviot Milksheep Speckled 2 Cadzow 2 (1960) Halfbred 128 BACKGROUND INFORMATION

Country Composite No. of Country Composite No. of breed foundation breed foundation population breedsa population breedsa Suffolk 2 (1810) MARC 3 (1984) Texel-Oxford 2 (1970) composite dam Wealdon Four- 6 (1971) line 2 quarter MARC terminal 3 (1986) Welsh Bleu 4 (1990)b sire composite Welsh Halfbred 2 (1955)b Minnesota 100 3 (1941) Welsh Masham 2 Minnesota 102 4 (1949) Welsh Mule 2 (1979)b Minnesota 105 3 (1949) Welsh 2 Morlam 8 (1961) Oldenbred Multinipple 4 (1923) Wensleydale 2 (1876) Polypay 4 (1979)b White Face 2 (1951)b Romeldale 2 (1915) Dartmoor Santa Cruz 3 (1920) 2 (1923)b Island USA California Red 2 (1971) Targhee 3 (1951)b Columbia 2 (1942)b Thribble Cross 3 (1903) Columbia- 2 (1943) Warhill 5 Southdale Willamette 3 (1952) Combo-6 6 (1970) Uruguay Merilín 2 (1910) Debouillet 2 (1954) Uzbekistan Akhangaran 3 Fannin sheep 2 Mutton-Wool Katahdin 3 (1957) Uzbek Mutton- 3 (1955) Montadale 2 (1933) Wool MARC 3 (1984) Yugoslavia Birka 2 composite dam Zimbabwe Wiltiper 2 (1946) line 1 a Year of origin or year recognised. b Year breeds society, association or stud book was established.

Source: Shrestha (2005)

A 5: ICAR Report of the working group on milk recording scheme

All tables, synthesising the results of the survey, are available on the ICAR website www.icar.org. BACKGROUND INFORMATION 129

Size of population of dairy sheep, type of milk recording and number of recorded ewes in ICAR member countries and other countries (year 2003) COUNTRY SIZE OF OFFICIAL MILK RECORDING POPULATION Total Number Of Number of Number of Average size of Females recorded females recorded flocks the recorded ( % population) flocks Belgium Few 277 5 55 Canada - - - Croatia 33,000 2,920 66 44 (8.8 %) Czech Rep. 473 204 8 26 (43 %) Denmark --- England & Wales Few 692 1 692 Estonia - - - France (1) 1,395,000 305,143 831 367 (22 %) Germany 20,000 1,193 11 (6.0%) Greece 8,732,000 7,885 102 77 (0.1 %) Israel (2) 46,200 14,975 22 681 (32 %) Italy 6,150,000 478,992 2,898 165 (7.8 %) Luxemburg - - - Mexico --- Morocco - - - The Netherlands 4,000 1,156 165 (29%) Norway --- Portugal (3) 110,000 24,242 (22 %) Slovak Rep 216,000 17,846 108 165 (8.3 %) Slovenia 6,300 1,704 32 53 (27 %) South Africa - - - Spain 2,361,000 188,197 400 470 (8.0 %) Switzerland 8,000 2,800 180 16 (35 %) Tunisia 25,000 2,393 9 266 (9.6 %) (1) In addition to A recording, France implements D method (616,337 females in 1,625 flocks, representing 44% of the whole dairy sheep population.

(2) figures from the survey 2002.

(3) Serra da Estrella and Mondegueira breeds only.

Source: Astruc et al. (2004) 130 BACKGROUND INFORMATION

Size of population and importance of milk recording for dairy sheep breeds accounting for more than 10 000 recorded ewes

BREED (country) SIZE OF OFFICIAL MILK OTHER MILK POPULATION RECORDING RECORDING (A or B or E nomenclature) (D nomenclature) Total number of Number of Number of Number of Number of females recorded recorded recorded recorded females (% flocks females (% flocks population) population) Assaf & Awassi (Israel) 46,200 14,975 22 (32 %)

Castellana (Spain) 250,000 11,852 14 (4.7 %) Churra (Spain) 650,000 31,868 78 (4.9 %) Comisana (Italy) 750,000 (*) 83,749 697 (11 %)

Corse (France) 100,000 19,607 70 (20 %) Lacaune (France) 825,000 179,299 399 584,507 1,503 (22 %) (71 %) Latxa Black-Faced 263,000 61,532 147 (Spain) (23 %) Latxa Blond-Faced 180,000 30,391 60 (Spain) (17 %) Basco-Bearnaise 80,000 18,738 82 3,557 19 (France) (23 %) (4.4%) Manech Tete Noire 120,000 17,675 355 1,559 6 (France) (15 %) (1.3%) Manech Tete Rousse 270,000 69,824 210 26,714 99 (France) (26 %) (10%) Manchega (Spain) 1,000,000 49,298 89 (4.9 %) (Italy) 207,000 (*) 62,756 362 (30 %) Sarda (Italy) 4,700,000 (*) 230,895 1,202 (4.5 %) Serra Da Estrella 105,000 21,000 (Portugal) (20 %) Valle Del Belice (Italy) 77,628 587

(*) figures from the survey 2000

Source: Astruc et al. (2004) BACKGROUND INFORMATION 131

Breeding scheme using artificial insemination (AI) Country AI Number of AI Selection Breed year of AI per year progeny-tested Criteria (***) starting (semen) rams per year France: Lacaune 1968 154,000 (fresh) * 458 FY+1.85 (281,000 out of the PY+0.2 P% nucleus) Manech Tete Rousse 1977 35,500 (fresh) * 130 FY+1.85 PY Manech Tete Noire 1977 (22,600 out of the 31 FY+1.85 PY nucleus) Basco-Bearnaise 1977 7,800 (fresh) * 39 FY+1.85 PY (2,000 out of the nucleus) 8,500 (fresh) * Corse 1992 (3,000 out of the nucleus) 20 MY 7,800 (fresh) Greece No Information For 2003 Israel (**): Assaf & Awassi 1999 2,000 (fresh) none Italy: Sarda 1986 19,256 (fresh)50 MY Portugal: Serra Da Estrella 1995 2,000 (fresh) 1 MY Spain: Latxa Blond-Faced 1984 9,997 (fresh) 36 MY Latxa Black-Faced 1985 18,730 (fresh) 61 MY Karranzana 1985 556 (fresh)7 MY Manchega 1988 30,166 (fresh)100 MY Churra 1986 1,500 (fresh) 60 MY + P% 11,500 (frozen) Castellana 1999 140 (fresh) 6 MY 562 (frozen) (*) in official milk recording (**) These are figures of the last survey in 2001 (***) MY = Milk Yield, FY = Fat Yield, PY = Protein Yield, F% = Fat Content, P% = Protein Content, SCC = Somatic Cell Count

Source: Astruc et al. (2004) 132 BACKGROUND INFORMATION

A 6: Changes in sheep breed composition in the USA

Annual registrations (head) 1980 1990 2000 Classification Barbados Blackbelly Watch Black Welsh Mountain Recovering Blue Face Leicester 50 70 Border Leicester 273 513 Rare Cheviot 2,518 2,851 2,136 Clun Forest Recovering Columbia 10,044 7,828 4,117 Corriedale 6,534 4,332 2,491 Cotswold 338 499 Rare Delaine Merinos 196 732 497 Dorper 2,165 Dorset 15,206 19,531 11,636 Dorset Horn Watch Finnsheep 1,069 1,034 264 Gulf Coast Native Critical Hampshire 21,360 16,460 10,018 Hog Island Critical Jacob 625 172 Rare Karakul 183 316 Rare Katahdin Hair Sheep 825 Recovering Lincoln 275 716 915 Watch Montadale 3,139 3,754 2,806 Natural Coloured 1,547 2,425 Navajo Churro 14 197 Rare North Country Cheviot 705 685 544 Oxford 1,457 1,914 1,593 Watch Polypay 11,874 1,935 Ramboullet 11,872 17,100 5,062 Romney 1,093 2,806 1,822 St. Croix Rare Santa Cruz Critical 250 219 Shetland 385 1,700 Recovering Shopshire 4,453 3,114 2,554 Recovering Soay Study Southdown 4,371 5,899 5,497 Recovering Suffolk 83,409 58,928 18,293 Targhee 1,445 2,538 1,570 Texel 347 Tunis 301 394 711 Rare Wiltshire Horn Rare Source: Country Report United States of America (2003) BACKGROUND INFORMATION 133

A 7: Country Report excerpts for sheep

Africa The most common exotic sheep breeds used in Botswana are Dorper, Karakul and Damara. The Dorper is often crossbred with the local Tswana for meat production. Under local conditions the Dorper is outperformed by the Tswana. The Karakul is mainly kept in south-western part of the country and is used for pelt production. At times farmers crossbreed Karakul with some meat-type sheep when the prices of pelt are not attractive. The Damara is mainly found in the northwestern part of the country predominantly among the Herero people. The Black-Headed Persian and the Ile-de-France are other sheep breeds, which are found in the country but are very few in numbers. In Ethiopia exotic sheep breeds used are Awassi, Hamshire, Bleu-de-Main, Merino, Romney, Corriedale and Dorper. Crossbreds of the local Menz breed with Awassi, Hamshire, Bleu-de-Main, Romney and Dorper are being tested for development and research. Uganda introduced exotic breeds like the Corriedale, Romney Marsh and Merino but they had no significant impact. The Merino and Doper sheep are now reared on a few farms in the country. The latter are kept mainly to meet the demand for lamb in urban areas. The Dorper is the predominant sheep breed in Zambian commercial production of lamb and mutton. Other important breeds are the Wiltiper, the South African Mutton Merino, the Blackhead Persian, the Dorset Horn, the van Rooy, the Corriedale and the Suffolk. Latin America In El Salvador the Criollo, Nubian, Saanen, Toggenburg and Franco Alpine exist, the Criollo being the most widely used. Additionally crosses of Criollo and Nubian are used. The Criollo origin is Spain. It first came into the country between 1548 and 1812. The Black Belly of Barbados originally came from Africa, and has adapted well to climate and utilisation. North America The sheep industry of Canada uses the Suffolk, Dorset, Kalahdin and others in farm flocks, and the Columbia in range flocks, although many of these are not registered. There are more than 40 breeds of sheep in Canada, but many are held in small numbers. The USA traditionally used Suffolk and Hamshire rams crossed with Rambouillet ewes to produce slaughter lambs in extensive production systems. In the mixed crop-livestock production system a wide variety of breeds are used including Suffolk, Hamshire, Rambouillet, Columbia and Dorset. Due to market forces there is growing interest in hair sheep breeds such as Dorper, Katahdin, St. Croix and Barbados Blackbelly, which do not need shearing. Currently Dorper (from South Africa) and Katahdin (a composite developed in the USA) are undergoing evaluation by researchers and industry and appear to be generating producer interest. Western and Northern Europe Ireland reports an inventory as summarised in the table below, with reference to origin and date of importation. 134 BACKGROUND INFORMATION

Origin of sheep imports to Ireland by breed

Breed Origin Imported in No. registered females Beltex Belgium 1996 272 Berrichon du Cher France 1993 120 Bleu du Maine Britain and France 1988 and 1989 300 Bluefaced Leicester Britain 1983 130 Charmoise France 1993 150 Charolais Britain and France 1984 and 1985 4,700 Ile de France Britain and France 1978 and 1987 150 Rouge de l’Ouest Britain and France 1988 650 Scottish Blackface Britain 19th century 1,000,000 Suffolk Britain 1891 13,192 Texel Netherlands and 1966 and 1992 5,000 France Vendeen France and Britain 1980 and 1985 860

Source: Country report Ireland (2003)

Latvia keeps a few Romanov, Gotland, Leicester, German Black-Headed sheep, Ile-de- France sheep but their existence is at risk. Sheep breeding farms deal with purebreeding and also with single blood addition of unrelated breeds to improve meat production. Farms dealing with meat production rear sheep of unknown origin. Sheep production is practiced only on a very modest scale in Sweden and exchanges with other countries are limited. In the 1980s, a certain amount of breeding material from dairy sheep breeds was imported into Sweden. This led to the establishment of some 20 dairy sheep herds for the production and direct sale of sheep cheese. In the United Kingdom only two exotic sheep breeds (British Texel and Charollais) play a major role in the sheep industry. Eastern and Southern Europe In Albania a transformation of local small ruminant breeds began 1955 to 1960, after importing Kaukaz Merino and Dagestan Tsygaia from the ex-Soviet Union. In the past Bosnia-Herzegovina tried many projects to improve production of sheep meat, milk and wool, both by selection among local breeds and crossbreeding with exotic breeds. The first project dates back to 1878 when 56 Karakul rams were imported from Bukhara, Uzbekistan. In general, all these attempts have failed. A comprehensive “Merinization” programme to improve wool production was carried out from 1947 to 1950. Local and imported Merino rams were used for AI of local ewes (approximately 800,000 ewes were inseminated in one year in the former Yugoslavia). As a result the first generation produced 25 to 50% more wool, but the sheep were more susceptible to diseases. In addition, farmers did not accept the programme, which was implemented without their BACKGROUND INFORMATION 135 consensus or participation. Local sheep genetic resources are almost totally preserved in their original form. However, sheep production will certainly be affected by socio- economic changes such as rural exodus, a need to increase efficiency and a ban of the nomadic system. These changes may also have some bearing on the genetic resources and cause further decline in the sheep numbers in the country. In Romania most extinct breeds were imported, not native. Friesian sheep were imported in large numbers to improve milk yield, but they disappeared. Many imported breeds were not adapted to the local breeding systems and to local parasites. Exotic sheep breeds have had an influence on the 32 breeds currently used in the Russian Federation. The Australian Merino and American Rambouillet have been crossed with local sheep to develop fine-fleece breeds. Romney Marsh sheep were developed into the dual-purpose Kuibeshevskaya breed in the mid 20th century. In 1978 the Russian long wool breed was created by crossing local sheep with Lincoln rams. The Soviet meat-wool breed was officially approved in 1986. The breed results from a cross between sheep of Northern-Caucasian meat-wool breed with Lincoln rams imported from Britain and Argentine with subsequent selection and interbreeding. In the last years of the 20th century meat breeds have been developed by crossing local ewes with Texel rams. Russia imported 226 female and 50 male Australian Merino and Texel from Australia and Finland 1990-1995. Further 500 Mongolian Sheep were imported to Russia 2001-2002 from China. Russia exported 193 male Romanovskaya, Tsigayskaya and, Orenburgskaya in 2001-2002 to Latvia, Kyrgyzstan and Kazakhstan. In Serbia and Montenegro Merino, Merinolandschaf (Württemberg), Ile de France, Bergamo and Jezersko-solčavska are used as exotic sheep breeds. In Slovenia only the Texel sheep is used as an exotic breed. Its population is increasing. Western Asia Armenia has successfully developed well adjusted wool breeds from a local rough-wool breed called Balbas with American Rambouillet and English Lincoln. Southern and Eastern Asia In Bhutan Merino and Comeback are the two exotic breeds of sheep used to cross with local sheep to improve the quality of wool and growth rate. Merino was fazed out as its wool attracted dust and it was difficult to process. The pure Comeback rams have been supplied to the farmers since early 1990s. The 9th plan livestock breeding strategy emphasises the production of adapted breeds by incorporating local sheep in the breeding programme. As an effect of imports and crossbreeding with exotic sheep, like the Merino in the earlier years and Comeback in the recent past, the numbers of local sheep and especially the black sheep population has decreased. China has listed the following exotic breeds used in the country: Charolais, Tsigai, Corriedale, Lincoln, Australian Merino, Romney, German Mutton Merino, Suffolk, Poll Dorset and Texel. Indonesia utilises both native and imported sheep. Sheep genetic material is imported both by government and traders. The government imported Romney and Kashmir in 1912. The introduction of the Kaapstad breed from South Africa is not clear. 136 BACKGROUND INFORMATION

No native sheep has ever existed in Japan. Sheep were imported after the Meiji era (1912) to produce meat and wool. In the early period, Corriedale was the main breed for wool production but in recent years, Suffolk is the main breed being utilised for meat production. Other breeds used are Southdown, Romney Marsh and Border Leicester. The sheep number has declined sharply because of the liberalisation of imports for sheep meat (1959) and wool (1961), along with the proliferation of synthetic fibres. Trotter breeds from Russia and England were imported into Kyrgyzstan for crossing with local sheep breeds. Rams of Australian Merino have been delivered to Kyrgyzstan since 1971 and widely used in crossbreeding. In 1998, 290 rams and 400 ewes of Australian Merino were delivered from Australia. They proved well adapted to local conditions and had a higher reproductive rate than local breeds. Among introduced sheep breeds of Nepal, the Polwarth is a continually introduced breed that is being used to improve wool production. Merino, Rambouillet, Border Leicester and Scottish Blackface are other introduced breeds. In Pakistan almost all sheep breeds are kept for wool and mutton. In the 1950s Rambouillet was imported for the first time and crossbred with several local breeds to improve wool quality and body weight. On government farms the following crossbreds are maintained: Pak-Awassi (Kachhi x Awassi), Lohi x Awassi, Pak-Karakul (Karakul x Kachhi), Salt Range x Afghani and Salt Range x Awassi. In the Philippines native sheep are believed to have originated from the Merino breed imported during the Spanish colonisation. Some other exotic breeds have recently been introduced primarily from the USA and Australia. These include Poll Dorset, Rambouillet Merino, St. Croix and Katahdin (found in selected government stock farms), Barbados Blackbelly (found in some commercial or institutional farms), Border Leicester, and Suffolk (in very small populations). The exotic breeds are also used for crossing and upgrading. The Philippines imported 119 sheep in 1990, 78 in 1991, 412 in 1992, 15 in 1994 and 678 in 1997. Many fine-fleeced sheep breeds like Caucasian, Vortemberian, Kirghiz, Australian Merinos and others were imported by Professor Lebedev to Tajikistan. In 1930 about 400 thousand head of imported sheep had already spread in the area between from Uzbekistan to Khatlon. Tajikistan imported the Darvaz breed since 1940, to improve productivity.

A 8: Country report excerpts for goats

Africa The Tswana is the major goats breeds used in Botswana. Other goat breeds of less significance in terms of numbers include the Saanen, Savanna Kalahari Red and Toggenburg. Many farmers are increasingly using the Boer goat for crossbreeding or keeping it as a purebred. The problem posed by the use of the Boer goat, especially in the eastern part of the country is its susceptibility to heart water. Under local conditions, the indigenous Tswana goat outperforms Boer goat in terms of productivity index. Almost all goats raised in Burundi are local breeds. Crossbreeding with the Alpine breed began in 1979 in provinces Ngozi and Kayanza to produce ½ and ¾ crossbreds for the BACKGROUND INFORMATION 137 rural areas. Ethiopia introduced exotic goat breeds to improve milk production. Anglo Nubian and Toggenburg were introduced by FARM-Africa and higher learning institutions. Crossbreds between Anglo Nubian and locals are being used for milk production by smallholders in central, eastern, south-eastern, and southern parts of the country. Toggenburg and their crosses with Hararghe Highland are used for research purposes at the Alemaya University of Agriculture and Debub University. Uganda has a small number of introduced goat breeds and their crosses. They include the Toggenburg, Anglo Nubian and the Saanen breeds. A crossbreeding programme with indigenous breeds to enhance milk and meat yield began in the early 1960s but was later abandoned following the period of civil strife of the 1970s and early 1980s. In 1995 the Boer goat from South Africa was introduced for crossbreeding with the locally adapted breeds. Currently, importation of Boer goats continues, mainly by Government, under the Strategic Interventions Programme. The breed is being multiplied at the National Animal Genetic Resources Centre farms and ranches as well as other privately owned farms for meat production. The demand by farmers for Boer goat is on increase; farmers and NGOs have continuously imported the breed into the country. This trend is likely to lead to significant changes in ratios of goat breeds in the country. North America The Canadian goat industry is relatively small, and at least ten breeds are present; Alpine, Nubian, and Saanen are the most important breeds. There is a system in place for genetic improvement for dairy purposes. The goat meat industry has been increasing with the introduction of Boer goats. Compared to other ruminant species, the number of goat breeds utilised in the USA is small. Angora goats are kept for Mohair. For twenty years Angora producers have participated in buck testing programmes where mohair quantity and quality are evaluated as well as body weight and feed utilisation. Several key lines of Angora have been developed in Texas. Periodically these producers have acquired additional genetic resources from South Africa. In 1992 the wool and mohair subsidy was removed. This has led to a decline in sheep and Angora goat numbers with a shift in breed choice toward hair sheep breeds. Six breeds of dairy goats are recognised (Nubian, Alpine, Saanen, Toggenburg, LaMancha and Oberhasli). The Nubian is also used for meat production due to its large body size. Specialised breeds for meat goat production are: the Spanish, Tennessee Stiff-legged (or Myotonic), and Boer. In the meat goat industry there has been a continuous conversion to the Boer goat, at the expense of the Spanish and the Tennessee Stiff-legged breed as well as the Angora. The original importations of the Boer goat were from New Zealand, due to the phytosanitary issues involved with direct importations from South Africa and the high importation costs. Positively, this importation created a number of opportunities for South African and USA breeders to exchange information and Boer germplasm. A flock of feral goats exists on San Clemente Island. Across the USA there are small populations of exotic breeds: Pygmy, Pygora, Nigerian Dwarf and the Kinder (note by the editor: Kinder is a cross between Nubian and Pygmy with about 1000 registered animals).

138 BACKGROUND INFORMATION

Eastern and Southern Europe Goat breeding is not a major sector in Russia. Between 1996-2000, about 310 female Lithuanian black head, Alpine, Nubian and Saanen goats were imported from Lithuania, Netherlands, USA, Denmark and Poland. Slovenia uses the Boer goat as the only exotic breed. The population is increasing. Southern and Eastern Asia China uses the Saanen and Boer goat as the only mentioned exotic goat breeds. Indonesia imported Ettawah goats from India. Saanen came from Australia in 1978. The way the Costa goat in Banten was brought into the country is not known. Angora or Montgomery goats were imported for the experimental stations in Bogor, Bandung and Padang Mangatas. The Dutch Improved Goat was brought to Java and Sumba. Boer goats, numbering less than 1,000 head, are distributed in North Sumatra, Lampung and South Sulawesi. They were imported from Australia since 1997 in order to upgrade local breeds. In Japan, apart from Japanese Saanen, which accounts for 30.4% of the total 14,000 goats, all other goats are of unidentified genotype and most of them seem to be crossbreds. Among introduced goat breeds of Nepal Jamnapari and Barbari are most commonly used introduced breeds. Other introduced breeds are Beetal, Ajmeri, Saanen and Kiko. Exotic breeds in the Philippines include the Anglo Nubian, Boer, Saanen, French Alpine and Toggenburg, all of which are found in commercial and institutional farms, and the LaMancha, available in very limited population. Crosses and upgrades of native goats are also widely available in the country. The number of goats imported in 1993, 1994, 1995, 1997 and 2001 were 40, 8, 16, 1,364 and 155. There is no information on the breeds imported.. From personal communication with Dr. A.D.N. Chandrasiri it was learned that Sri Lanka imported 200 Beetal goats from Pakistan in 1990 and 1991. In 1992, 28 German Fawn goats were imported from Germany followed by 22 Boer goats in 1996. In 1997, 368 Jamnapari came from India. All exotics are used for breed improvement. Tajikistan imported Angora goats in 1937 and created a new fibre breed by crossbreeding with local goats. Saanen, Alpine and Boer (from the USA) were introduced to Vietnam in recent years for improving milk and meat production with the aim of adaptation and crossing with local breeds. Additionally, Barbari, Jamnapari and Beetal have been introduced since 1994. They are used for crossing with the Co or Bach thao goats to improve production. Oceania Palau only has Anglo Nubian goats, which were introduced from the Northern Mariana Islands since 1983. They do not make significant contribution to production. The numbers are declining and the breed is expected to disappear over the next decade.

BACKGROUND INFORMATION 139

A 9: Possible introduction routes of cattle from West to South, Southeast, East and Northeast Asia, with centres of domestication

Source: Payne and Hodges (1997)

A 10: Possible introduction routes of cattle into North, West and north-eastern Africa

Source: Payne and Hodges (1997) 140 BACKGROUND INFORMATION

A 11: Possible introduction routes of cattle into the Indian subcontinent and Southeast Asia and probable centres of domestication for Bos (bibos) spp.

Source: Payne and Hodges (1997) BACKGROUND INFORMATION 141

A 12: Cattle migration after the 16th century

(a) Spaniards take cattle to America in the 16th century, (b) Dutch, English and French colonisers take cattle to North America in the 17th century, (c) British cattle to Australia at the end of 18th century, (d) Indian Zebu cattle to America and Australia in the mid 19th century, (e) constant movement of cattle throughout Maghreb

Source: adapted from Jasiorowski et al. (1988)

A 13: Total Simmental population per country** in the early 1990s

Germany Yugoslavia France* Austria Hungary South Africa USA Switzerland Italy England Sweden Ireland

012345 number of animals in 1'000'000

* including Montbeliarde, Simmental Francais and Abondance; ** only the countries with the largest Simmental populations

Source: own elaboration, data from Künzi and Stranzinger (1993) 142 BACKGROUND INFORMATION

A 14: German Simmental export to Turkey

25,000 23,133

20,000

15,000

10,000

numberanimals of 7,831

5,054 4,851 5,000 2,880 2,611 2,310 2,110 717 92 0 205 450 00 0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Source: own elaboration, data from Averdunk et al. (2001)

A 15: EU heifer export by origin and year

1995 1996 1997 1998 1999 2000 2001 2002 2003 Total Germany 67,731 65,490 69,056 78,226 89,796 70,457 31,134 40,121 27,881 539,892 Netherlands 37,175 42,389 36,297 35,302 28,426 29,314 9,199 20,491 17,996 256,589 France 33,969 36,690 19,878 27,647 27,799 24,153 25,975 34,170 25,916 256,197 Austria 22,367 28,492 22,396 14,528 17,398 12,170 18,194 9,154 11,063 155,762 Denmark 11,062 10,705 8,198 11,187 11,941 6,346 4,337 7,059 8,652 79,487 Belgium- 2,473 3,503 3,882 2,992 3,487 2,761 7,815 9,824 3,700 40,437 Luxemburg Spain 1,512 760 991 373 190 3,386 889 9,933 2,208 20,242 Ireland 4,245 112 39 323 257 1,641 831 1,111 2,059 10,618 Italy 1,213 901 1,068 878 814 1,031 823 594 641 7,963 Sweden 178 229 1 381 54 40 86 490 906 2,365 Finland 4 220 2 10 0 42 30 0 0 308 Greece 40 0 179 0 0 0 0 0 0 219 UK 22 0 0 0 0 0 0 0 0 22 Portugal 0 10 0 0 0 0 0 0 0 10 Total 181,991 189,501 161,987 171,847 180,162 151,341 99,313 132,947 101,022 1,370,111

Source: Eurostat (2004) BACKGROUND INFORMATION 143

A 16: Annual EU heifer export by destination 1993-2003

WESTERN A SIA TOTA L 4 Mor oc c o 13 , 4 4 6 Algeria 7,040 Tunisia 2,406 Egy pt 2,209 Libyan Arab Jamahiriya 1, 3 18 Sudan 25 Turkey 15, 2 3 7 Lebanon 2,137 Jordan 1, 8 0 6 U.A . Emirates 1, 2 8 2 Saudi Arabia 797 Kuw ait 476 Qatar 12 3 Syria 53 Oman 49 Yemen 28 Israel 24 Bahrain 21 Gaza + Jericho 16 Georgia 14 A SIA TOTA L 434 Philippines 231 Thailand 1 Turkmenistan 61 Uzbekistan 37 Kasakhstan 33 India 24 Iran, Sri Lanka, Bangladesh, Reunion 12 North Korea 14 South Korea 12 Japan 5 Hong Kong 4 A FRICA TOTA L 228 Rw anda 51 Malaw i 31 Ethiopia 11 Senegal 36 Mauritania 19 AfricaNigeria, Niger, Asia Ivory Coast, Sierra Western Asia/Maghreb Leone 21 Uganda 47 South Africa 11 Cameroon 2 M. S. W. E. E. SC. SE. W. Asia Maghreb not specified 45 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,00 number of heifers

Source: Eurostat (2004) 144 BACKGROUND INFORMATION

A 17: Annual EU heifer import and export 1998-2002

400,000 900,000 836,632 337,615 exports imports net-import exports 300,000 675,000 imports net export net export net-import

200,000 165,660 450,000 140,728

82,507 100,000 225,000 49,522 16,979 27,041 4,781 00 03,538 82 6,222 1,957 0 0 -3,632 -46 -5,463 -20 -2,166 -9,578 -1,978 -11,081 -9,784 -14,469 -10,907 -36,052 -23,812 number of animals number of animals -100,000 -225,000

-135,257

-200,000 -450,000

-568,768 -300,000 -675,000 -304,523

-400,000 -900,000 UK Italy Spain Ireland France Austria Greece Finland total EU total Sweden Belgium Portugal Denmark Germany Belg.-Luxbg Netherlands net-importer net-exporter

Source: Eurostat (2004)

A 18: EU heifer import by destination and year

250,000

200,000

Others 150,000

Ireland

100,000 Portugal number of animals

Germany 50,000 Italy

0 Spain 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Source: Eurostat (2004) BACKGROUND INFORMATION 145

A 19: Annual EU heifer import by origin 1993-2003

number of animals

0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000 100,000

intra-EU 9 6,759 extra-EU 14 , 3 8 2 Bulgaria 90 Czech Rep. 8,560 Hungary 3,147 Latvia 4 Poland 396 Rep.Macedonia 3 Romania 40 Slovakia 403 Sw itzerland 1, 16 8 Ic eland 0 28

NorthNorw East ay Canada 474 N.

Am. Europe USA 8 Algeria 45 Hong Kong 4

Others Mor oc c o 9 not specified 2

0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 10,000 number of animals

Source: Eurostat (2004)

A 20: German Simmental export by year

35,000

30,000

25,000

20,000

15,000 number of animals 10,000

5,000

0 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Source: own elaboration, data from Arbeitsgemeinschaft Deutscher Rinderzüchter e.V. (2002) 146 BACKGROUND INFORMATION

A 21: Introduction of the Simmental breed to different environments

Aspect Breed Place Success Source Adaptation to Simmental Harsh Superior production Speers, 1997; extensive ranch environments parameters Schoeman, 1996 conditions Tick infestation Simmental Southern Higher management Rechav et al., under extensive Africa requirements than 1991 ranch conditions local breeds Production potential Simmental x Venezuela Highest potential Plasse et al., 1995 Brahman among different crosses but only with good management Farm reconstruction Simmental, Kosovo, Farmers more Cossee, 2003; Project after Balkan Brown Swiss Bosnia- satisfied with World Bank, crisis Herzegovina Simmental than with 2003; IFAD, Brown Swiss 2004 Intensive production Simmental Denmark High daily weight Dansk systems gains Simmental, 2004; Hansen, 2004 Peri-urban Simmental, Iran Important Schahidi et al., production systems Brown Swiss, contribution to milk 2001 (no forage Jersey, Danish production but production) Red and dependent on Holstein subsidies

BACKGROUND INFORMATION 147

A 22: Crosses of Simmental with local cattle breeds

Crossing partner Place Time Source Unnamed local breeds Iran 1940s Schahidi et al., 2001 Botswana 1990s Nsoso and Morake, 1999 Mozambique 1980s Rocha et al., 1987 South Africa 1970s Borstlap, 1972; Neser et al., 2002 China 1980s Liang, 1988 Mexico 1990s Country Report of Mexico (2003) Iran 1990s Schahidi et al., 2001 Turkey 1990s Oklahoma State University, 2004a Colombia 1990s Orbita, 2004 Boran Kenya 1970s Sonn, 1985 Boran Ethiopia 1970s - 1990s Demeke et al., 2003 Barca Ethiopia 1970s - 1990s Demeke et al., 2003 Horro Ethiopia 1970s - 1990s Demeke et al., 2003 Southern Anatolian Turkey 1990s Ertugrul et al., 1999 Red Holstein crossbreds Indonesia 1990s Hadi et al., 2002

A 23: German Simmental export to Balkan countries 1998-2002

10,000 29 Serbia and Montenegro

9,000 Kosovo

8,000 67 Croatia 4,997 7,000 401 Bosnia 6,000

332 5,000 80 799 4,000

number of animals 6,917 3,000 4,725 2,000 3,909

1,000 525 1,116 295 96 0 62 1998 1999 2000 2001 2002

Source: own elaboration, data from Arbeitsgemeinschaft Süddeutscher Rinderzucht und Besamungsstationen e.V. (2002) 148 BACKGROUND INFORMATION

A 24: Semen export of industrialised countries by origin in 1991

USA N.

Am. Canada France 700 Germany 569 Netherlands 507 Belgium 285 Denmark 85 United Kingdom 62 Spain 11 Ireland 0 Es tonia 625 Hungary 17 Slovenia 16 Europe EU Eastern Poland 3 Austria 74 Sw itzerland 36 Finland 33 Sw eden 23 Europe Western Norw ay 10 New Zealand 12 0 Japan 34 South Africa 14 Others Australia 7 0 250 500 750 1,000 number of doses in thousands

Source: own elaboration, data from Chupin and Schuh (1993)

A 25: EU semen export by destination in 2003 (%)

Middle East and North Other Sub Sahara Africa Africa 13% Eastern Europe except 17% CIS 5% CIS 4% North America Other 4% Latin America 26% Oceania 17% 4% Far East 7% South Africa Other Western Europe 1% and accession countries EU 12% 16%

Source: Eurostat (2004) BACKGROUND INFORMATION 149

A 26: EU semen export by destination in 2003 (number of doses)

Utd. Kingdom 490,462 Sweden 117,422 Spain 519,252 Portugal 198,038 Netherlands 485,650 Luxembourg 58,025 Italy 452,218 Ireland 104,387 Greece 50,600 France 465,751 Fr Germany 254,055 Finland 83,523 Denmark 51,905 Belgium 193,016 Austria 187,503 Switzerland 183,671 Slovenia 3,850 Slovakia 50,045 Poland 78,566 Norway 375,244 Malta 2,500 Lithuania 73,132 Latvia 16,445 AccessionHungary EU 87,968 Estonia 32,952

Other W Europe and Czech Rep. 180,085 Cyprus 3,454 Serbia and Montenegro 21,008 Romania 7,432 Croatia 21,320 CIS 42,855

except Bosnia-Herz. Europe Eastern Albania 3,527 Ukraine 15,550 Kasakhstan 17,078 Belarus 1,200 Azerbaijan* 1 Greenland 597 Faroe Isles 600 Canada 73,963 North 302,836 America CIS USA New Zealand 38,134 N. Caledonia 68 Fr.Polynesia 800 182,553 Oceania Australia 0 100,000 200,000 300,000 400,000 500,000 600,000

150 BACKGROUND INFORMATION

Venezuela* 0 Uruguay 5,915 Surinam* 1 Peru 1,901 Nicaragua 400 Mexico 208,448 Guatemala 1,000 El Salvador 16,690 Ecuador 37,382 Cuba 1,090 Costa Rica 17,825 Colombia 97,946 Chile 70,709 Brazil 138,076 Bolivia 8,000 Argentina 20,000 Turkey 207,693 Tunisia 124,560 Syria 13,000 Sudan 2,000 Saudi Arabia 6,610 Qatar* 2 Oman 552 Morocco 31,428 Lebanon 13,273 Kuwait* 1 Jordan 50 Israel 18,451 Iran 1,000 Egypt 28,011 Bahrain 3,552 Algeria 12,100 Thailand* 1 Taiwan 250 South Korea* 3 Philippines* 4 North Korea 1,000 India 20,802 Hong Kong* 0 South Africa 38,060 Zimbabwe 800 Zambia* 3 Tanzania 7,679 Saint Helena 140 Nigeria 150 Mauritius 2,500 Malawi* 0 Ivory Coast 1,300 Guinea 500 Ethiopia 4,000 Cameroon 450 Burundi* 0 Other Sub Sahara Africa Far EastBotswana Middle East and North Africa Latin America 320 0 100,000 200,000 300,000 400,000 500,000 600,000

* only number of consignments given, total number of doses exported in 2003: 6,579,839 Source: Eurostat (2004) BACKGROUND INFORMATION 151

A 27: Semen import of developed countries by origin in 1991

Luxemburg* Italy* Germany United Kingdom France 430 483 Spain 247 Netherlands 205 Ireland 168 Belgium 158 Portugal 155 Greece 105 Denmark 50 Switzerland 318 Austria 204 Sweden 45 Finland 12 Malta 6 Norway 5 Australia 220 Japan 154 South Africa 111 Hungary 144 Czech Republic 138 Poland 47 Romania 32 Slovakia 30 Estonia 24 Macedonia 10 Slovenia 2 Croatia 1 USA* N. Canada 287 Amer. Europe E. Europe OthersW. EU 0 100 200 300 400 500 number of doses in 1,000 *no data

Source: own elaboration, data from Chupin and Thibier (1995) 152 BACKGROUND INFORMATION

A 28: EU semen import by origin in 2003

Partner Trade Quantity (doses) % Total 7,531,240 100.0 Intra-EU 2,929,304 38.9 Extra-EU 4,601,936 61.1 USA 2,476,008 32.9 Canada 1,759,377 23.4 Czech Republic 92,720 1.2 Hungary 79,441 1.1 New Zealand 68,286 0.9 Slovakia 42,040 0.6 Australia 28,090 0.4 Switzerland 25,982 0.3 Poland 18,974 0.3 Romania 3,700 0.0 Slovenia 3,500 0.0 Taiwan 2,000 0.0 Norway 501 0.0 Colombia 500 0.0 Ivory Coast 400 0.0 Ukraine 400 0.0 Japan 10* 0.0 Egypt 7* 0.0 Uruguay 0* 0.0 * number of consignments Source: Eurostat (2004) BACKGROUND INFORMATION 153

A 29: Semen imports and exports by breed group and region in 1991

3,500,000 3,166,000 3,000,000 2,500,000 2,000,000 1,500,000 956,000 1,000,000 665,000 637,300

500,000 29,000 206,000 63,000 173 , 9 9 3 57,9 02 33,000 70 0 18,910 94,053 0 00 0 -15,000 -55,6 00 -170,000 -117,000 -152,244 -236,500-135,300 -51,239 -153,548 -114,762 -500,000 -384,855 -599,817 -1,000,000 numberof semen doses -1,500,000 -1,665,900 -2,000,000 import export beef beef beef beef beef dairy dairy dairy dairy dairy dual dual dual dual dual breeds breeds breeds breeds breeds breeds breeds breeds breeds breeds purpose purpose purpose purpose purpose Canada EU (8) Eastern Europe Western Europe other developed (12) (7) (4)

number of included countries in paranthesis

Source: own elaboration, data from Chupin and Thibier (1995)

A 30: Semen imports to developing countries in 1991 by breed, breed group and destination

Latin America (n=28)

tropical breeds Asia (n=23) beef breeds

other dairy breeds Near East (n=16) Jersey

Holstein Africa (n=37)

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 number of imported semen doses in 1'000'000

Source: own elaboration, data from Chupin and Schuh (1993) 154 BACKGROUND INFORMATION

A 31: Annual semen imports and exports of EU in 2002/2003

2,000,000 exp ort s imports 8,000,000

1, 5 5 4 , 9 151, 4 5 1, 9 85,9 8 55,58 6 1,500,000 6,000,000 net expo rt net -imp o rt 1,066,094 1,000,000 4,000,000

601,518 500,000 368,952 2,000,000 274,342323,404 12 5 , 14 1 69,239 82,212 31,474 0 560 0 5,750 0 0 -3,086 -88,025-64,161-18,656 -92,445 -207,058 -131,266 -283,520-252,061 -500,000 -2,000,000 -498,395 -630,906 -603,905 -797,858 -1,000,000 -4,000,000

-1,272,970 -1,500,000 -1,438,189 -6,000,000 -6,382,497

-2,000,000 -8,000,000

net -impo rt er net -expo rt er

Source: Eurostat (2004)

A 32: German semen import (doses) by year and origin

1,400,000

1,200,000

1,000,000

800,000 Others North-America 600,000 EU

400,000

200,000

0 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Source: own elaboration, data from Arbeitsgemeinschaft Deutscher Rinderzüchter e.V. (2002) BACKGROUND INFORMATION 155

A 33: US Brahman export by destination and year

Head 1977 1978 1979 1980 1981 1982 Mexico 27 46129 258 839 334 Ecuador 8 125 4 141 South Africa 9 4 14 46 42 Dominican Republic 3 33 21 72 40 Panama 45 44 31 171 86 36 Guatemala 176 134143 16 20 31 Venezuela 431 1177 1 43 30 Costa Rica 26 31 45 48 76 24 Philippines 6 14 64 1 16 Colombia 204 396279 214 130 13 total 932 786 760 744 1,316 707

Source: own elaboration, data from Cowert (1983)

A 34: Brazilian Simmental embryo import by origin 1986-1993

Canada 1,096

Germany 472

Italy 160

Argentina 78

Sw itzerland 20

USA 9

0 200 400 600 800 1,000 1,200 number of embryos

Source: own elaboration, data from Fraga (2004) 156 BACKGROUND INFORMATION

A 35: Brazilian Simmental semen import by origin 1972-1993

Germany 78,595

USA 45,258

Canada 15,582

Italy 14,400

Switzerland 9,630

Argentina 1,500

Austria 1,100

0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000

number of semen doses

Source: own elaboration, data from Fraga (2004)

A 36: Brazilian semen doses sold in 2002 by breed

Nellore 3.24

Angus 1.34

Holstein 1.17

Milking Gyr 0.33

Jersey 0.30

Simmental 0.27

Guzaret 0.22

Brown Swiss 0.05

other 0.08

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 semen doses in million

Source: own elaboration, data from Country report Brazil (2003) BACKGROUND INFORMATION 157

A 37: Israeli Holstein export by country and time period

Iran 1960-1979 20,485 2,175 Egypt 1981-1984 1,750 1,468 Zambia 1960-1980 530 518 Hungary 1989-1990 457 407 Bulgaria 1996 320 180 Turkey 1986 103 93 India 1996 43

0 5,000 10,000 15,000 20,000 25,000 number of animals

Source: own elaboration, data from Israeli Cattle Breeder’s Association (2004)

A 38: AI coverage by breed and country groups for developed countries in 1991

100.0% dual-purpose breeds beef breeds 83.8% 77.7% 76 .2% dairy breeds 73.8 % 75.0% 72 .9%

63.6% 64.0% 57.6 %

49.0% 50.0%

33.4% 30.4%

25.0% 14 . 5 %

7.8 % 4.5% 5.2 %

0.0% Eastern Europe European Union North America Western Europe Other developed All developed (n=12) (n=12) (n=2) (n=7) (n=4)

Source: own elaboration, data from Chupin and Thibier (1995) 158 BACKGROUND INFORMATION

A 39: AI coverage by breed and country groups for developing countries in 1991

100.0% local breeds crossbreds 77.4% temperate breeds 75.0%

60.0%

53.1% 51.9% 50 .7% 50.0%

25.0% 23.3% 17 . 9 % 14 . 5 % 13 . 2 %

7.0% 6.2% 5.1% 2.9% 1. 7% 0.1% 0.0% Africa (n=37) Asia (n=23) Latin America Near East (n=16) All developing (n=28)

Source: own elaboration, data from Chupin and Schuh (1993)

A 40: Use of AI by breed group and region for developed countries in 1991

100%

76% 64% 75% 64%

dairy breeds beef breeds 84% 58% 74% 8% dual-purpose breeds 50% 49%

30%

25% 73% 78% 33% 15% 5% 5% 0% 0% 0% 0% Eastern European North Western Other All Europe Union (n=7) America Europe developed developed (n=8) (n=2) (n=3) (n=4)

Source: own elaboration, data from Chupin and Thibier (1995) BACKGROUND INFORMATION 159

A 41: Use of AI by breed group and region for developing countries in 1991

100%

17% 27% 35% 34% 10% 75% 60% temperate breeds

30% crossbreds local breeds 50% 31%

73% 59%

25% 28% 43% 35%

12% 7% 0% Africa (n=5) Asia (n=8) Latin America Near East All developing (n=5) (n=5)

Source: own elaboration, data from Chupin and Schuh (1993)

A 42: Share of bull breeds used in Botswana from 1987 to 1995

South Devon Pinzgauer Sussex Tswana Tuli 0.2% 0.4% 0.6% 0.8% 1.2% Africander 0.1% Friesian 2.6% Bonsmara Hereford 3.4% 0.2% Santa Gertrudis 5.1%

Charolais 13.2%

Brahman 56.5%

Simmental 15.7%

Source: own elaboration, data from Nsoso and Morake (1999) 160 BACKGROUND INFORMATION

A 43: Australian beef cattle registrations by breed 1990 1995 2000 British breeds and Australian derivates Angus 13,280 16,623 22,611 Poll Hereford 33,763 26,941 20,277 Hereford 32,764 24,067 17,637 Murray Grey 11,072 11,680 8,761 Shorthorn 8,119 6,350 6,718 South Devon 1,863 1,266 1,635 285 1,755 1,575 Red Poll 2,077 908 788 Devon 2,190 1,242 710 Australian Lowline n.a. n.a. 461 714 451 299 Galloway 559 328 179 British White n.a. n.a. 93 Group total 106,686 91,611 81,744 Tropical breeds Brahman 18,496 13,427 11,932 Santa Gertrudis 9,477 8,180 8,312 Droughtmaster 4,044 6,007 6,793 Brangus 1,068 1,432 1,842 Braford 4,176 2,263 1,518 Belmont Red 1,033 704 1,476 Charbray 493 245 350 Tuli -60 161 Boran -41 26 Sahiwal 299 158 n.a. Greyman -71 n.a. Group total 39,086 32,588 32,410 European breeds and derivates Limousin 9,403 10,407 4,479 Charolais 3,939 4,152 4,055 Simmental 13,669 7,823 3,861 Gelbvieh 187 1,044 1,023 Romagnola 562 977 573 Blonde d’Aquitaine 599 1,055 457 Salers 1,637 870 363 Maine Anjou 503 561 134 Original and Australian Braunvieh 200 121 127 Association Piemontese -161 80 Belgian Blue n.a. 198 47 Chiangus 103 21 36 Chianina 168 113 34 Group total 30,970 27,503 15,269 Other Wagyu n.a. 300 1,260 Mandalong Specials n.a. 388 86 Grand total 176,742 151,702 129,423 Source: own elaboration, data from Allen (2002) BACKGROUND INFORMATION 161

A 44: Number of breeders by breeds in South Africa in July 2003

19 others Charolais Brangus Sussex Drakensberg Afrikaner Hereford Angus Santa Gertudis Simbra Limousin Nguni Bonsmara Simmental Brahman

0 50 100 150 200 250 300 350 400 450 500

number of breeders

Source: own elaboration, data from Simmental South Africa (2004)

A 45: Ratio of utilisation of the Simmental breed in different countries for milk and beef

milk : meat France 60-80 : 20-40 Hungary 75 : 25 Austria 60 : 40 Switzerland 60 : 40 Yugoslavia 55 : 45 Germany 50 : 50 Italy 50 : 50 South Africa 5 : 95 England 1 : 99 Sweden 1 : 99 Ireland 0 : 100 USA 0 : 100

Source: Künzi and Stranzinger (1993)

162 BACKGROUND INFORMATION

A 46: Examples of absorption of local cattle by imported breeds

Absorbed Absorbing Place Time Aspects Source breed breed Criollo European Argentina Late 19th New breeds displaced Giovambattista breeds and and early Criollo Cattle to marginal et al., 2001 Zebu cattle 20th century regions breeds Criollo European Latin Late 19th Introduction of exotic Alba, 1978 and Zebu America and early breeds led to the cattle 20th century replacement of Criollo breeds cattle Sahiwal Temperate Pakistan Second half Number of purebred Philipsson, dairy 20th century Sahiwals was drastically 2002 breeds reduced by crossbreeding Kenana Holstein Sudan Since 1980s Efforts to increase milk Philipsson, yields to improve milk 2002 supplies Muturu, White West Since 1970s Changing ecological Jabbar and Keteku, Fulani Africa conditions and human Diedhiou, N’Dama demographic pressures 2003 Biu Zebuine West Since 1970s Changing socio- Blench, 1999 cattle Africa economic framework conditions Rahaji Azawak North- 1990s Better environmental Blench, 1999 west adaptation Nigeria Two Black and Lithuania 1990s Development accelerated Maleviciute et native White and by change of economic al., 2002 cattle Red cattle framework conditions breeds

A 47: Holstein’s and Sahiwal’s contribution to composite breeds

Composite breed Contributing breeds Place Source with Holstein besides Holstein contribution Karan Fries 50-62% Tharparkar, India Felius, 1995 Sahiwal Frieswal Sahiwal India Mudgal and Arora, 1994 Drakensberger Africander South Africa Oklahoma State University, 2004b Siboney 37% Zebu Cuba Robes and Martinez, 1990 Mambi 25% Zebu Cuba Robes and Martinez, 1990 Tropical Holstein 3% Zebu Cuba Robes and Martinez, BACKGROUND INFORMATION 163

1990 Lucerna Shorthorn, Criollo Colombia Blake, 2004 Simhol Simmental Colombia Orbita, 2004 Holgus Angus Texas/USA Felius, 1995 Taino, Mestizo Holstein, Girolando, Mantiqueira, Riopardenze, Felius, 1995 Guzerando, Brazilian Dairy Hybrid, Tropical, Hays Converter, RX3, Beef Machine, Ranger

Composite breed Contributing breeds Place Source with Sahiwal besides Sahiwal contribution Australian Milking Jersey, Red Sindhi Australia Felius, 1995 Zebu Australian Friesian Holstein Australia Felius, 1995 Sahiwal Jamaica Hope Jersey; Holstein Jamaica Trail and Gregory, 1981 Taurindicus Holstein New Zealand Felius, 1995 Frieswal Holstein India Mudgal and Arora, 1994 Pabna Milking Cow Hariana, Red Sindhi Bangladesh Islam and Bhuiyan, 1997 Karan-Fries, Karan- Felius, 1995 Swiss, Quasah, Mpwapwa

A 48: Performance of the Holstein breed in different environments

Aspect Place Performance Source Coat colour Cuba Black cows have higher milk yields Gutierrez et al., 1985 Economic returns to Venezuela Positive economic returns only if Holmann et al., AI high daughter milk response to sire 1991 selection is given Comparison to Venezuela Profitability of purebred Holsteins Holmann et al., crossbreds in dual- lower if comprehensive economic 1990a purpose systems analysis is carried out Genotype by Colombia, Smaller daughter milk responses in Stanton et al., environment Mexico, Puerto Latin America compared to USA 1991 interactions Rico Investment in semen Colombia, Economic returns from investments Holmann et al., from North America Mexico, in semen from North America are 1990b Venezuela negative in most scenarios Conception rate Mexico Lower conception rates than Brown Osorio and Swiss, Zebu and the respective Segura, 2002 crosses 164 BACKGROUND INFORMATION

Aspect Place Performance Source Performance Mexico Factors affecting performance are McDowell et affecting factors basically the same as in temperate al., 1976 regions Sire comparisons Mexico Correlations between sire McDowell et comparison values in Mexico and al., 1976 North America are high Heat tolerance Indonesia Better adaptation to high altitudes but Pangestu et al., still not reach their genetic potential 2000 Replacement rate India High involuntary and low Mangurkar et reproductive performance limit al., 1986 replacement rate Crosses in co- Bangladesh Holstein x local compared to local Hoque et al., operative setting and Sahiwal x local has highest 1999 performance Heat tolerance Kenya Severe heat stress in lowlands but in Ikiror, 2001 highlands only transient Smallholder farms Uganda Several management constraints Nassuns- show low adaptability to smallholder Musoke, 2002 conditions Introduction in Ghana Feeding and management practices Gyawu et al., humid forest zone need to be improved to increase 1988 performance parameters Intensive production North Africa Under intensive management Ageeb and systems comparable production parameters Hayes, 2000; Sadek et al., 1994; Sonn, 1984 Import of heifers Morocco Under intensive management Johnson et al., from US comparable production parameters 1989 Economic Morocco High economic sensitivity due to Srairi and El performance climatic conditions Khattabi, 2001

A 49: Country Report excerpts for cattle

Africa In southern Africa, Botswana and Zimbabwe reported breed substitution through the use of exotic breeds. Between 1987 and 1995 the use of exotic bull breeds in Botswana was significantly higher than that of indigenous bull breeds both in natural service (94.9 vs. 5.1% per year) and as semen (94.1 vs. 5.9% per year). The most significant breed in southern Africa is the Brahman. In West Africa, Burkina Faso reports great influence of exotic genes in bovines. The Azaouak Zebu, which is widely spread in the country for its recognised dairy performance, is noted. The Red Bororo of Niger and the Goudali of Nigeria are recently introduced breeds valued for their dairy and beef performance. The introduction of these breeds was driven by pastoralists and transhumants on the one hand and peri-urban private producers BACKGROUND INFORMATION 165 on the other. The Tarentaise, Brown Swiss and Montbéliard breeds are consistently imported for improvement of dairy production, as are the Jersey, Holstein and Limousin breeds. Burundi reports the introduction of Sahiwal for crossbreeding with Ankole since 1953, followed by introduction of European exotics since 1970. Exotic breeds are mainly used in dairy production in Burundi as well as in Ghana. Ghana imported Friesians, Jersey and Sahiwal from Europe and India to develop a milking animal. In Cameroon, several breeds are used either in purebreeding or for crossbreeding, namely the Holstein, Montbéliard, , Salers, Jersey, Charolais, Brahman and Baoulé breeds. Mali reports consistent importation of Montbéliard and Holstein cattle. Eritrea and Ethiopia in eastern Africa report import of exotic animals and semen of dairy cattle breeds like Holstein-Friesian, Jersey and Simmental for improvement of production. Eritrea specifies an import of Holstein-Friesian from Holland in 1991, and semen of unspecified breeds from Europe, Kenya and, more recently, from Ethiopia. In Malawi, locally adapted Brahman and consistently imported Fresian, Holstein, and Jersey cattle are used for crossbreeding in meat and milk production. South Africa is one of the main sources of exotic genetic material in the southern African region. Some of the exotic cattle breeds used in Zambia were introduced in colonial times. Settler farmers introduced exotic breeds into Zambia as early as 1945. Since then, exotic breeds were continuously imported into the country, first in live animals during the 1970s and later through imported semen from Europe and America. Many breeds continue to be consistently imported, including Friesian, Guernsey, Jersey and Simmental, Pinzgauer, Boran, Brahman or Hereford. Latin America El Salvador is the only South American country reporting gene flow. The first exotic breeds imported in 1890 were Durham or Shorthorn, Holstein and Normande. In 1923, these were followed by import of Holstein, Ayrshire, Jersey and other breeds. In the 1950s, Zebu from Guatemala and Guzerat and Santa Gertrudis from Texas were imported. Most recently, the American Brahman was introduced. The most important beef breed in El Salvador is the Brahman, while the most significant dairy breed is the Holstein. North America The Canadian dairy herd is made up of Holstein cows (95%), Ayrshires and Jerseys. The beef industry was based primarily on British breeds until the late 1960s, when many Continental European breeds were imported. Recently there has been a move back to smaller brood cows, in concert with the United States industry. Canada is a net-exporter of live and processed cattle. Much of the international trade in live animals takes place through transfer of genetic material. In 2001, 2,452,991 doses of semen and 2,337 embryos of both dairy and beef cattle were exported. The USA imported a large number of continental European breeds in the 1970s and, at various points in time, beef producers imported Bos indicus breeds. The USA Holstein has been developed as a global breed. The USA cooperates internationally on a number of issues related to animal genetic resources. With regard to the conservation of animal genetic resources, collaborative effort spans North America, South America, Europe, 166 BACKGROUND INFORMATION conservation programmes which are at a similar stage of development. Across the USA there are a number of federal and state scientists that interact with scientists of other nations on the use of animal genetic resources and mapping animal genomes. USA animal genetic resources are traded globally and this trade has led to increased animal performance and economic returns and reduced poverty. The USA Holstein has made substantial contributions to the global dairy industry. Western and Northern Europe Finland reports importation of farm animal genetic material with cost-effective long-term benefits. The imported volumes of breeds have been sufficient to start breeding programmes. Nevertheless, new foreign populations quickly exceeded the level of Finnish animals. For example, when the breeding of the Ayrshire and the Friesian outpaced the breeding of Finncattle, these breeds soon replaced Finncattle in dairy production. Dairy production is dominated by a few powerful Holstein breeding organisations. In Ireland the Shorthorn was crossed with British Friesian cattle during the 1940s and 1950s. During the late 1980s the Friesian cattle were crossed with higher-yielding Holstein-Friesian cattle from Continental Europe and North America. Today almost all dairy cattle in the country are Holstein Friesian type. Simmental cattle have been imported from Austria since 1971. Similarly in Sweden international exchange is intensive and widespread, and dairy cattle breeders regularly import breeding material. The Swedish breeding practices, which focus on health and functionality, are of interest to foreign livestock breeders. Semen doses of Swedish Red and White bulls have been exported to Argentina and New Zealand. In the United Kingdom the dairy industry is dominated by purebred Holstein and the beef industry is dominated by Limousin, Charolais and Belgian Blue cattle. Most native beef breeds have experienced introgression from exotic breeds. Eastern and Southern Europe In Albania the first crosses of local breeds and Jersey were produced in the 1930s. After 1965, Black and White heifers were imported from the Netherlands, Denmark and Germany, and a national AI system and bull centres with bulls of Jersey, Black and White and Simmental breeds were established. The number of the local cattle is decreasing due to transformation of local cattle breeds into "cultivated" ones by crossbreeding with imported breeds. Belarus reports the introduction of Holstein from European countries (Holland, Denmark, Germany) and the USA. Exotic breeds were first introduced to Bosnia-Herzegovina from Austria in 1886. In 1906, over 2,000 bulls were imported from Tyrol. A new local breed (Gatacko cattle) was created in the south of the country by crossing of Tyrolean Grey with local Busa cows. Simmental cattle were introduced to neighbouring Croatia and Serbia in the first decades of the 20th century, and from there imported to Bosnia-Herzegovina. In subsequent years, breeding animals came from Germany, Austria and other countries. In 1950, research on the breed composition of herds in the Sava valley (in northern Bosnia) showed that 13.7% of cows were of the local Busa breed, 7.8% were Busa type crosses, 23.8% were Posavsko cattle, 15.7% were Pinzgauer and Pinzgauer x Busa crosses, 7.0% were crosses between BACKGROUND INFORMATION 167 cattle in the country were of the Simmental breed and its crosses with other breeds, including the Busa. Friesian cattle were imported in 1960, followed by later imports of Holsteins. In 1996, a three-year programme to rehabilitate the animal production sector in the Federation of Bosnia and Herzegovina was adopted. It envisaged import of 60,000 high quality cows, 100,000 sheep and 20,000 goats. During the first year of the programme (1997) some 10,000 heifers were imported from Hungary, Austria, Germany and the Netherlands. 75% were Simmental, 10% were Friesian and Holstein, 10% were Montafon (Alpine Brown), and 5% were Oberinntal (Grey Tyrolean). Both pregnant heifers and doses of semen for use in AI have also been imported. Farmers wanting to purchase the imported animals received soft loans from the Government in accordance with criteria approved by the Federal Ministry of Agriculture, which required that over 50% of the production assets of the farm must have been destroyed and that land be available (one cow or 5 sheep per hectare of the agricultural land). In general, the policy was to give one cow to one family. At a later stage, the option for more market-oriented production units of 3-5 cows with an interest in increasing production and raising efficiency through improved management techniques prevailed. A similar programme for the republic of Srpska set a target of 20,000 heifers to be imported in a three-year period starting in 1997. About 2,000 heifers were imported in 1997 on an International Fund for Agricultural Development (IFAD) loan. The majority of the imported heifers were Simmental cattle and roughly 20% were Montafons, Tyroleans and Frisians. In Bosnia-Herzegovina the importation of exotic breeds increased production and led to the formation of a well-adapted new breed (Gatacko) and new, locally adapted strains of Tyrolean Grey, Montafon and Simmental. On the other hand, importation was not accompanied by changes in production systems and intensification. Many breeds have been experimented with in Bulgaria over the last 20 years. As a result, it has been established that the Hornless Hereford is the best breed for mountainous regions and the Simmental and its hybrid forms are best for plains. In the Czech Republic an originally unified Simmental-type red-pied cattle was developed. The current genetic diversity of farm animals in the Czech Republic reflects the previous orientation of production targets pursued in the process of breeding in different species and breeds. One example is the development of the originally unified Simmental- type red-pied cattle throughout the modern-day Czech Republic. Efforts to increase milk yields led to both the controlled immigration of Ayrshire and, later, Red-pied Holstein cattle in half of the population, and to the crossbreeding (later conversion-crossing) with Lowland black-pied and later Holstein cattle through the import of animals and sperm in the other half. This resulted in the concurrent realisation of separate selection programmes and the development of separate populations (breeds), namely the Bohemian red-pied and the Holstein. The gradual development of beef cattle farming led to breeding specialisation of a subset of the original Bohemian red-pied herd and to the emergence of a population of meat-type Simmental cattle which were developed and managed separately. In Hungary, replacement of local breeds with foreign cattle occurred in two stages. The first occurred at the end of the 19th Century and the first half of the 20th Century, when 168 BACKGROUND INFORMATION

(Hungarian Grey Cattle with the Simmental). In the second stage, the spread of technologically-advanced large-scale cattle farming throughout the country was accompanied by the import of modern, intensive, specialised varieties and hybrids suitable for large farms and by the crossing of local semi-intensive varieties. Market pressure to import new varieties, especially hybrids, is still present, but the importance of preserving endangered and native varieties is receiving increasing emphasis. The Russian Federation imported from Switzerland in 1861. Today many breeds are selected using Brown Swiss. A subsequent substantial import of genetic resources from Switzerland took place between 1920 and 1930. Between 1958 and 1972 imports came from Austria, Switzerland, Hungary and USA. After 1972, semen and sires were imported from USA and Canada. Since 1995, small herds of Angler, Red Dutch, Red Lithuanian, Red Estonian and Brown Latvian breeds have been combined with Red Steppe breed. The table below gives an overview of export and import of cattle into and out of the Russian Federation between 1990 and 2002.

Import and Export of cattle from Russia from 1990 to 2002

Breed, cross Year of Exporting country Number purchase of animals Import males females Dairy Ayrshire, Brown 1990 1995 Germany, Denmark, 273 15,655 Swiss, Holstein, Netherlands, USA, Red and White Finland, Japan Ayrshire, Angler, 1996 2000 Germany, Denmark, 379 10,376 Jersey, Brown Netherlands, Hungary, Swiss, Holstein, Finland, Estonia, Italy, Simmental, Red Canada and White, Black and White Ayrshire, Angler, 2001 2002 Austria, Germany, 174 14,141 Jersey, Brown Netherlands Denmark, Swiss, Holstein, France, Finland, Estonia, Simmental, Red Lithuania and White, Black and White, Red Danish, Red Steppe Beef Hereford 1990 1995 Canada 7 Hereford, 1996 2000 France, Canada, Ukraine, 22 1,195 Limousine, Aubrac, Germany, Hungary Salers, Charolais Aberdeen Angus, 2001 2002 France, Hungary, 93 1,670 Galloway, Germany Limousine, Aubrac, Salers, Charolais Export Dairy Ayrshire, Holstein, 1996 2000 Kazakhstan 10 250 Kostromskaya, BACKGROUND INFORMATION 169

Black and White 2001 2002 Kazakhstan, Tajikistan 24 1,321

Source: Country report Russian Federation (2003)

In Serbia and Montenegro, the exotic breeds used are: Simmental, Holstein-Friesian, Holstein (Black and Red), Limousin, Charolais and Montafon. The use of local breeds is decreasing. In Slovakia, a massive import of highly productive, specialised, foreign breeds is taking place. Some local breeds have become endangered, mainly due to reduction in biodiversity, economic pressure to increase production, and gradual market globalisation. Since 1989, global processes like holsteinization i.e. upgrading of local cattle breeds with the Holstein have greatly contributed to a decline in biodiversity in Slovakia. In Slovenia, there is widespread use of Simmental and Black and White Holstein cattle, both of which exist in stable populations. Other stable populations are the Charolais, Galloway and Red Angus, while populations of Limousin, Scottish , Red Holstein and Montbéliard are increasing. In the early 18th century, the first immigrants from Germany and Holland introduced Red cattle. Angler and Dutch breeds were crossed with the red cattle at the end of the 19th century. At the end of the 18th century, the first Simmental animals appeared in Russia. By 1950, the Sychevskaya breed - a cross between local cattle and Simmental - had been developed. Until the 1960s Simmental and related breeds dominated in Russia (95% of the total population). After 1990, the Simmental population began to decrease and the Red and white dairy breed - a cross between Red and white Holstein and Simmental - was approved in Russia. Western Asia In Armenia, Swiss breeds were selected to cross with Grey Caucasian Cattle because of similar climatic conditions in the mountainous region in Switzerland. In Azerbaijan efforts were made to raise milk productivity by bringing male cattle of the milk-trended breeds of “Qara-Ala” (black-black and white). While milk productivity increased, the milk quality indexes were reduced. The number of highly productive pedigree animals imported from abroad is decreasing each year due to lack of adaptation. In Turkey, the most widespread exotic breeds are Holstein Friesian, Brown Swiss and Turkish Brown Swiss. The Turkish Brown Swiss is a development of the Karacabey Brown, which originated from crosses of Brown Swiss with Anatolian Grey. Holstein and Jersey were successfully crossed with local breeds, while crossing with Angus and Hereford was unsuccessful. As a result, there is a loss of breeds in Turkey due to geography, socio-economic and cultural structure, and presence of livestock improvement projects. Southern and Eastern Asia Bhutan has applied crossbreeding programmes with the local Nublang (Siri) and Indian breeds such as Haryana, Sahiwal and Red Sindhi. In the early 1980s, AI of Nublang (Siri) with germplasm of Jersey and Brown Swiss was introduced. Jersey has been used throughout the country, while the Brown Swiss breed has been used only in the high altitude areas. Crossbreeding trials with Australian Milking Zebu and Tarentaise were not satisfactory. 170 BACKGROUND INFORMATION

According to incomplete statistics, China introduced more than 120 domestic animal breeds before 1989 from Russia, the UK, France, Germany, the Netherlands, Denmark, the USA, Canada, Australia and New Zealand, including 27 cattle breeds, 25 sheep breeds, 4 goat breeds, and 11 pig breeds. The imported domestic animals enriched genetic resources in China and played a positive role in improving local animal breeds. In the last few years, the annual value of breeding livestock imports exceeded US$ 20 million. Some of the exotic breeds used in China today are the Holstein, Piedmontese, Simmental, South Devon, Jersey, Limousin, Shorthorn, Angus, Hereford, Deutsches Gelbvieh, Charolais and Red Dane. The China Holstein and China Simmental breeds have also been developed. Indonesia imported Holstein Friesian first from West Friesland and then from Australia, New Zealand, the USA, Japan and Canada. The population of Holstein Friesian cattle in 2002 was roughly 354,000. Imported cattle are mainly used for feed-lot operators. 26.4% of cattle used in Japan for beef production are Holsteins and 31.3% are Holstein hybrids. The Holstein is similarly dominant in the dairy sector, comprising nearly 100% of dairy cattle since 1965. In Kyrgyzstan increased demand for livestock products and the development of a domestic processing industry have stimulated breeding throughout the livestock branch. Kyrgyzstan is developing composite breeds to create cultural and highly productive breeds. This has resulted in the creation of two new breeds. There are regular imports of live male breeding animals and frozen semen of various breeds for local breed improvement. The main countries of origin for cattle imports are Austria, the USA, Holland, Switzerland, Russia. In early 1984 in Laos, 120 pure Holstein-Friesian heifers were introduced from Cuba. The cows were not properly managed and could not survive in the difficult local conditions. Myanmar introduced a crossbreeding programme for dairy production in 1977, using fresh semen from imported dairy breeds such as the Friesian and Jersey. Jersey and Holstein-Friesian are the most commonly used breeds imported to Nepal. Other introduced breeds are Brown Swiss, Ayrshire, Sahiwal and Haryana. In Pakistan, dairy breeds such as Holstein Friesian, Jersey, Illawarra Shorthorn and Brown Swiss (recently imported by multinationals) have been imported for use in purebred herds and for crossing. Imported and locally produced Jersey semen has also been used in the irrigated regions and, more so, in the Barani areas. Holstein and Jersey semen is locally produced. Semen from breeds such as Australian Illawarra Shorthorn, Swedish Red and White, Chinese Black and White, Australian Friesian Sahiwal, Australian Milking Zebu and Brown Swiss has been imported for use in crossbreeding, but at a very limited extent. There has also been limited use of Charolais and Simmental in experimental purposes and in crossbreeding for beef. In the Philippines cattle of the Bos indicus line and Bos indicus x Bos taurus crosses are most common. In feed-lot fattening American and Australian Brahmans and their crosses are used. Imported Holstein Friesian cattle and their crosses are used as dairy herds. Breeding is conducted through AI using locally produced or imported frozen semen. The table below gives an overview of the numbers imported between 1990 and 2001. BACKGROUND INFORMATION 171

Animal numbers imported to the Philippines, 1990-2001

Year 1990 1991 1992 1993 1994 1995 Cattle 23,022 15,773 49,237 82,130 115,916 168,769 - Breeder 1,709 3,099 15,875 7,458 6,430 6,299 Year 1996 1997 1998 1999 2000 2001(P) Cattle 167,435 156,719 186,835 236,882 196,777 102,478 - Breeder 2,645 1,269 704 725 2,056 30

P - Preliminary

Source: Country report Philippines (2003)

In Sri Lanka Dr. A.D.N. Chandrasiri reports imports of semen between 1990 and 2001 in a personal communication. Details are provided in the table below.

Import of cattle semen to Sri Lanka by breed and country of origin

Breed of Semen No. of doses Country of Origin Jersey (all types) 136,110 New Zealand, Australia, Canada, Netherlands Denmark Friesian (all types) 62,500 Australia Canada Holland New Zealand Holstein Friesian 46,000 New Zealand, Australia, Canada, Netherlands Australian Friesian Sahiwal 14,000 Australia Australian Milking Zebu 6,025 New Zealand, Australia, Canada, Netherlands Ayrshire 3,500 New Zealand Sindhi 596 India Brown Swiss 270 Switzerland

Source: Country report Sri Lanka (2003)

Tajikistan imported exotic breeds, mainly from Russia, to improve productivity after 1940. Cattle breeds imported include Friesians, Swiss breeds and Kholmogosrkiy. Nevertheless, many imported breeds did not adapt to the local conditions and did not become wide-spread. In Uzbekistan it is reported that exotic, locally adapted breeds of cattle, namely Black and White, Alpine Brown, red-steppe, red Lithuanian, Latvian and red Estonian. Recently Holstein, kept only in several pedigree farms, and Santa Gertrudis have been introduced. In Vietnam a “Zebuization” programme is taking place since 1994, in which local cows are crossed with Red Sindhi or Ongole bulls to create the Lai-sind. For improved dairy production the Lai-sind is crossed with Holstein Friesian, and to improve beef production it is crossed with Charolais, Limousine, Hereford or Brahman. To improve draught power, Lai-sind is continually upgraded to Red Sindhi. 172 BACKGROUND INFORMATION

Locally produced cattle semen fills about 60-65% of the current demand for semen doses; the remainder is imported from France, Australia, the USA, Cuba and other countries. Imported breeds include the Red Sindhi, Sahiwal, Brahman, Charolais, Limousine, Hereford, Simmental, Santa Gertrudis, Droughtmaster, Belmont Red, Red Brangus Red Brahman, Jersey and Holstein Friesians. The Brahman is considered to be well adapted in the central region of Vietnam. The most common increasingly used breeds are the Holstein Friesian from Cuba, the USA and Australia, the Sindhi from Pakistan, and the Jersey from the USA. Oceania The first cattle were introduced to Samoa by European settlers in the late 1800s. The first breeds to be introduced were mainly temperate breeds such as the Aberdeen Angus, Ayrshire, Friesian, Hereford, Shorthorn and Red Poll from Australia and New Zealand. Subsequently, Brahman cattle were introduced. In 1984, embryo transfer technology was used to bring the Piedmontese breed from the USA. Breeds used today are Braford, Brahman, Droughtmaster and Santa Gertrudis, Australian Friesian Sahiwal, Friesian and Jersey.

A 50: Annual EU breeding pig import and export 1992-2003

60,000 exports imports 150,000 125,644 49,493 exports imports 40,000 net export net-import 100,000 net export net-import 22,112 15,996 20,000 12,219 50,000 7,565 8,689 5,018 2,547 1,396 74 7 320 118 90 0 0 -58 -99 -1,889 -1 -14 -3,935-2,351 -5,079 -7,981 -12,662 -20,000 -50,000 -20,459-17,484 -23,063

-40,000 -100,000 -38,983

-134,060 -60,000 -150,000 UK Italy Spain Ireland France Finland Austria Greece total EU total Portugal Sweden Denmark Germany Belg.-Luxbg Netherlands net-importer net-exporter

Source: Eurostat (2004); Warning: It is possible that these numbers are distorted due to misclassification of breeding pigs. BACKGROUND INFORMATION 173

A 51: Imported pig breeds during the 1970s into different countries

Importing country Belgian Belgian Landrace Belgian Belgian Landrace Belgian German BritishLandrace Landrace German East Landrace German Landrace Netherlands Australian Landrace Danish Landrace Finish Landrace Landrace Norwegian Landrace Swedish Landrace Thailand Pietrain Belgian Pietrain German Pietrain Other White Large British Welsh British WelshN.Ireland Australian Tamworth Yorkshire Netherlands Yorkshire Canadian US Yorkshire Hampshire Canadian US Hampshire Hampshire Other Duroc Canadian US Duroc Duroc Other Lacombe Canadian US Chester White China US Poland US Spot Austria x x x xx xx Belgium x x x Bulgaria x xxx Denmark x x x x x x x Finland x x former Czechoslovakia x x x x x x France xxx Germany xx xx x xx xx x xxx Greece x x x x Hungary xxxxxxx Europe Ireland x x x Italy xx Lithuania x x Netherlands x x x x x x Norway xx Poland x Portugal xx Sweden x x x x x x x Switzerland xx x UK x x x x x x x x x x Yugoslavia x x x x x x Japan xxx Asia Malaysia xxx Myanmar x x x Philippines x x x x x x x x x South Brazil x Ecuador xx America Argentina x x x Northern Canada xx x x x America USA xxxx x x x xxx Australia x x x x x x x Oceania New Zealand xx Papua New Guinea x x

Source: adapted from Sutherland et al. (1985) 174 BACKGROUND INFORMATION

A 52: State of the pig genetic resource utilisation in Uruguay

Breed Adaptation Population Use Importance Breeding Use of Production Breeders Improvement Character- Reproduction tendency intensity product system Association programmes isation Criollo locally decreasing low not defined mainly family extensive non- without phenotypic/ natural adapted/risk crossbreeding existent registration molecular of extinction Duroc continuous decreasing medium economic mainly industry/ semi- for breed genealogical phenotypic natural Jersey imports importance crossbreeding family intensive group register Hampshire locally decreasing low economic mainly industry/ semi- non- without phenotypic natural adapted importance crossbreeding family intensive existent registration Landrace continuous increasing high economic mainly industry semi- for breed genealogical phenotypic natural, imports importance crossbr., intensive group register artificial some purebr. insemination Belgian continuous stable low economic mainly industry intensive non- without phenotypic natural, Landrace imports importance crossbreeding existent registration artificial insemination Large continuous increasing high economic mainly industry semi- for breed genealogical phenotypic natural, White imports importance crossbr., intensive group register artificial some purebr. insemination Pampa locally stable low economic mainly family/ semi- for breed genealogical/ phenotypic/ natural adapted importance crossbr., industry intensive group productive molecular some purebr. register Spotted locally decreasing low economic mainly family/ semi- non- without phenotypic natural Poland adapted importance crossbreeding industry intensive existent registration Pietrain continuous stable low economic mainly industry intensive non- without phenotypic natural, imports importance crossbreeding existent registration artificial insemination Source: Country Report Uruguay (2003) BACKGROUND INFORMATION 175

A 53: Country Report excerpts for pigs

Africa In Burundi local pig breeds are crossed with exotic breeds, mainly Large White, in order to raise productivity. Mali reports constant imports of Large white, Yorkshire and Chinese pig breeds. The exotic pig breeds in Uganda include the large white, , and the Landrace. The major pig breeds of Zambia’s commercial sector are the exotic Landrace, the Large White, the Duroc and the recently introduced Dalland. Latin America El Salvador uses the following pig breeds in modernised systems: Landrace, Yorkshire, Duroc and crosses between them. In the rural zones local breeds dominate. North America In Canada pigs of seven different breeds were registered in 2001, but 99.4% of these were in the Yorkshire, Landrace, and Duroc breeds, which are used predominantly by the pig industry either pure or in crossbreeding programmes. These breeds, along with newly developed lines, continue to be used in crossbreeding programmes, often without registering individuals. Composite lines are playing a larger role as breeding animals in the commercial industry of the USA. The most prominent sire breeds are Duroc, Hampshire and Berkshire, while Yorkshire and Landrace are maternal breeds. Western and Northern Europe In Denmark the British breeding company PIC has established breeding and production herds, and is responsible for about 10% of supplies of the breeding animals and boar semen. Owing to the strict provisions governing the importation of genetic material, Iceland has not participated in the globalisation, which has characterised cattle breeding in various parts of the world for over a decade. However, breeding activities involving pigs, poultry and furred animals are primarily non-Icelandic activities, as the breeding stock in these sectors is regularly and frequently imported. Genetic material in the pig sector is based on regular imports of overseas breeds which are used in systematic crossbreeding, as is the case in other countries and therefore there are no independent domestic breeding programmes. Purebreeding and systematic breed crossing are used in pig breeding in Latvia. Nucleus farms produce gilts for industrial crossing. 9.2 % of sows are reared on nucleus farms from locally adapted breeds (Latvian White improved with Yorkshire and Landrace). Recently introduced breeds (Yorkshire and Duroc) are renewed with imported representatives. Only Hampshire and Pietrain boars are imported continuously. In the past, the Dutch pig breeding organisations imported various foreign breeds for their 176 BACKGROUND INFORMATION expected since no material that would contribute positively to commercial pig breeding programmes is available outside breeding organisations. Pig breeding in Sweden is currently the subject of internal debate, and opportunities to co- operate with Danish breeders are under inquiry. The native Swedish breeds have satisfactorily met breeding goals, such as good fertility and high growth rates. In the early years of the last century, breeding pigs were imported mainly from Denmark in order to strengthen native Swedish breeds. Importing breeding material from native breeds into Sweden continued with the introduction of breeding boars and semen from Norway and Finland. From the mid-1950s breeders have used Yorkshire breeding pigs imported from Canada and Finland in addition to native Swedish stock. The pig industry of United Kingdom is focused on producing hybrids based upon lines or breeds of global importance. Systematic crossbreeding, within-breed strain crosses and hybridisation is widely practised in the pig and poultry industries. 79% of purchased boars (of which 47% are Large White) are supplied by four companies, and 70% of purchased hybrid gilts by three companies. Eastern and Southern Europe In Belarus the Belorussian Black pig breed was created by difficult reproductive crossbreeding of local pigs with European pedigree breeds: Big Yorkshire, Medium White, Dark-nap, Large Black, Berkshire, Long-eared Lincoln. Certain groups of boars are imported to Poland every year, in particular Landrace and Large White, which are included in the breeding of the Polish Landrace and Polish Large White breeds. These imports are from England, France and the Scandinavian countries. Boars are also imported systematically for sire breeds, though this is limited and insufficient to improve the leanness considerably. Other exotic breeds mentioned are Belgian Landrace, Duroc, Hampshire, Pietrain and Synthetic Line 990. The pig sector of the Russian Federation is represented mainly by Large White, which accounts for 88% of the total pig stock. The Large white breed was first imported from Britain in the 1980s to be crossed with local breeds. Later, in the 20th century, genetic stock of Large White breeds were imported again from Britain. As a result, the population of Soviet Large White breed was established that proved to be highly adaptive to different climatic and natural conditions of the former USSR. This Large White breed is superior to other breeds in many breeding indices, therefore it takes the leading position in Russia’s pig sector. Examples of synthetic lines developed in the Russian Federation are listed in the table below.

Development of synthetic pig breeds in the Russian Federation

Breed name Type Registration Breeds used /Recognition year Breight general 1948 Large White, Danish Landrace, Lithuanian and purpose Latvian breeds and Polesskaya breed from Belorussia Northern 1955 local breeds with Large White, Berkshire and Caucasus White short ear breeds BACKGROUND INFORMATION 177

Muromskaya 1957 local breeds with Lithuanian White and Large White Kemerovskaya general 1961 native Siberian pigs, Large White and purpose Berkshire Belorussian black 1976 in 19th century with local breeds with and white Yorkshire, Large Black and other imported breeds; expanded in the 1920s by using Swedish Landrace, Estonian Beacon, Large White and Berkshire breeds CM-1 early 1990s Large White and others maturing Livenskaya general unknown local pigs with Large White and Berkshire purpose Tsyvilskaya unknown native Chuvash pigs with Large White

Source: Country report Russian Federation (2003)

The exotic pig breeds used in Serbia and Montenegro are Swedish Landrace, Large White, Dutch Landrace, German Landrace, Belgian Landrace, Danish Landrace, Canadian Landrace, Pietrain, Hampshire and Duroc. Pig breeds in Slovenia, are classified into locally adapted pig breeds and exotic pig breeds and the trends in population size is estimated. The exotic breeds that are widely used are Pietrain and Duroc. Another exotic breed is Large White. Locally adapted breeds are Swedish Landrace, German Landrace and Large White. Western Asia Armenia imported the breed in 1930. Initially they were used to improve local pig characteristics, and later for purebreeding and crossbreeding. The main breeds cultivated in Armenia are local meat, big white, exotic Welsh, Landrace, Duroc and Ukrainian lines. Populations in north-eastern regions have developed a local breed well adapted to local conditions. Southern and Eastern Asia In Bhutan Large Black, Duroc Jersey, Yorkshire, Saddle Back and Landrace are the exotic breeds used to cross with local pigs for improving body weight, litter size and growth rate. As the preference by farmers is for coloured breeds, the nucleus farms will maintain only Duroc Jersey, Saddle Back, Large Black and Hampshire. In addition, there are also plans to introduce Meishan pigs as a maternal line to improve the litter size of the crossbreds. Experiments using the exotic and local breeds are developing a composite breed appropriate for village management conditions. Crossbreeding with exotic breeds like Large White, Yorkshire, Duroc, Large black, Saddle black and Hampshire has been promoted by the Government. As a result, the ratio of indigenous pigs fell by about 47% between the years 1987 and 2000. Plans are under way to revive nucleus breeding farms for the native pigs. The pigs in the farm will be used for crossbreeding with some exotic lines to upgrade its inferior trait in fertility and litter size which will then be supplied to the farmers as per their choice. Large-scale pig production was introduced to Cambodia in 1997. Two products are piglets 178 BACKGROUND INFORMATION for sale and meat. Breeds are imported from the surrounding countries Thailand, Vietnam, and Taiwan China reports importing the following exotic breeds of pigs: Berkshire, Landrace, Yorkshire, Duroc, Soviet White, Hampshire, Kemiroff, Pietrain. After 1960 Indonesia imported Hampshire, Duroc, Landrace, Poland China, and Yorkshire (Large white) pigs. The potential area for pigs is in North Sumatra, West Kalimantan. Before 1960 there were German Landrace, Netherlands’s Landrace, Tamworth, Saddleback pigs available in Indonesia. Pig production in Japan is characterised by the high popularity of the Kagoshima Black Pig. In addition to this breed, there are hybrids from triple crossbreeding Large White, Landrace, and Duroc, and hybrids from triple crossbreeding Large White, Landrace, Berkshire (in place of Duroc), or commercial pigs which have been produced using pigs imported from foreign breeding companies. Eight breeds comprising , Berkshire, Landrace, Large White, Hampshire, Duroc, Spot, and Chester White have been included in the statistics. In 1999, Duroc, which is used as a terminal sire in a triple cross, accounted for by far the highest percentage at 52.9% of the total of 77,000 breeding boars; Berkshire, Landrace, Large White accounted for 6.1%, 5.3%, and 6.6% respectively. Other than these breeds, exotic hybrid pigs and crossbreds for triple cross account for 13.9%, and 13.4% respectively, making up more than 99% of the pigs in this category (see the table below). On the other hand, of the total 886,000 breeding sows, Berkshire, Landrace, and Large White account only for 3.7%, 4.2%, and 3.1% respectively, with Duroc at 1.8%. Exotic hybrid pigs and crossbreds for triple cross account for 14.0% and 72.5% respectively. Purebred female account for only 13.5% of the total. Pigs used for meat production total 5.97 million, only 6.6% of which are used as purebreds. Berkshire at 2.8%, Duroc at 1.3%, Landrace at 1.1%, and Large White at 0.8% are used for this purpose. Berkshire, which is extremely popular in Japan under the name Black Pig, accounts for about 50%. Kagoshima Black Pig accounts for only 2% (327,000 animals) of the annual total number of slaughtered pigs (16 million, FY 2001).

Transition in numbers of breeding hogs, breeding sows, and pigs for fattening in Japan

1. Breeding 1975 1980 1985 1990 1995 1999 hogs Middle 368 72 55 63 62 75 White Berkshire 316 310 879 1.685 2.236 4.717 Landrace 11.693 10.876 7.108 6.209 4.421 4.120 Large 3.600 7.926 9.632 9.316 5.395 5.147 White Hampshire 13.406 21.660 11.099 5.077 1.774 1.009 Duroc 3.052 12.775 38.672 48.391 42.962 40.934 Spotted 63 403 200 41 24 4 Chester 12 40 52 67 20 19 White BACKGROUND INFORMATION 179

Others 23 284 3.303 9.441 8.898 11.047 Hybrids 3.303 8.336 8.690 10.769 Cross 931 2.279 3.134 9.485 9.087 10.356 Total 33.464 56.625 74.134 89.775 74.879 77.428 2. Breeding sows Middle 8.754 1.803 327 215 115 172 White Berkshire 11.551 6.551 5.509 9.861 15.490 32.950 Landrace 392.318 321.305 154.783 75.480 42.367 37.387 Large 38.878 54.357 50.457 54.610 32.325 27.053 White Hampshire 52.050 55.960 13.743 5.630 2.572 1.203 Duroc 9.754 41.246 41.034 26.259 14.320 15.977 Spotted 903 894 286 59 38 0 Chester 75 176 1.313 220 40 105 White Others 1.145 409 40.742 135.421 165.686 128.517 Hybrids 40.742 110.672 128.839 124.361 Cross 295.085 663.341 841.611 796.346 606.019 642.693 Total 810.513 1.146.042 1.149.805 1.104.101 878.972 886.057

Source: Country report Japan (2003)

In these excellent Japanese statistics the indication of the breed composition of “others, hybrids and crosses”, which constitute a major part (more than 100%?) of the sows and 40% of the boars, is missing. In Kyrgyzstan’s pig breeding “Big White Breed”, North Caucasian, Landrace, German Yorkshire were imported for domestic breeding. In Malaysia the poultry and pig sub-sectors are very dependent on continually imported breeding stock and feedstuffs. Practically all production activities for poultry and pigs are based on the use of exotic breeds. This is because there are no local poultry or pig breeds suitable for modern, commercial production. Indigenous pig breeds of Myanmar in various regions are slow in growth, thick fat and of low productivity. At present, the Government Enterprise and private companies import high productive breeds to crossbreed with local breeds. That’s why the pig population has increased in recent years. In some areas commercial boars are kept for breeding. Most of the boars are crossbred with exotic breeds like Large White, Landrace, Duroc Jersey and Berkshire. Myanmar had imported some exotic breeds since 25 years ago for improving pig breeding. Apart from native Nepalese pig breeds there are Large White Yorkshire, Landrace, Hampshire, Saddleback, Fayuen, Tamworth and Duroc pigs in Nepal. Among these 180 BACKGROUND INFORMATION

Yorkshire and Landrace are most widely used. The Philippines recognises pigs as the single largest contributor to the local livestock sector. Its pig population is made up of several breeds and genetic groups that are distributed throughout the country and in different farming production systems. The pig genetic groups present in the country are classified into: exotic standard purebreds, synthetic hybrids and the Philippine native pig and its upgrades or crosses. The purebred group is composed mainly of the Landrace, Large White, Duroc and Pietrain breeds. Berkshire and Hampshire breeds are also present but in significantly lesser numbers compared to the four aforementioned breeds. The Landrace and the Large White are extensively used by commercial pig farms for the production of a crossbred sow line, while the Duroc and Pietrain are commercially used for boar line. Synthetic hybrids like the New Dalland, Babcock, Seghers, Cotswold and Hypor are also continuously introduced into the country and commercially utilised by many commercial pig raisers. Recent importation of these exotic breeds and hybrids came from Canada, USA, European countries and Australia. Sri Lanka imported mainly Large White, Landrace and Duroc pigs for breed improvement (see the table below).

Pig imports to Sri Lanka by year and country of origin

Year Country Breed Number of No. of semen animals doses 1990 Australia Landrace/ Large 120 White 1991 - - 2 1996 USA - 32 1997 USA Large White 17 USA Land Race 8 USA Duroc 6 1998 USA Land Race 15 USA Large White 13 Unknown USA Duroc 100 USA Land Race 45 USA Large White 45

Source: Country report Sri Lanka (2003)

Imported breeds play an important role in Vietnam’s pig production (see the table below). 56% of intensive farms and only 9% small households raise exotic breeds, while crossbreds are found mainly in small households (58%) and are less common in intensive farms (28%). Regarding breed supply, only 5% of the breeds in small households are provided by state intensive farms. The rest are raised or traded by small households themselves, which led to a poor quality of breeds. The semen for pig breeding is internally produced. High BACKGROUND INFORMATION 181 productivity exotic breeds will be used to meet the requirements of urban consumption and the export market. In domestic consumption and in rural areas usually hybrids are used. Most Vietnamese slaughter pigs are commercial F1 crosses between local sows (mainly Mong Cai), and exotic boars such as Duroc, Landrace, Yorkshire or Hampshire breeds. Some exotic breeds such as Large White, Hampshire, Berkshire and German Yorkshire have decreased in numbers due to degradation of genetic breeding over many years importation and/or not meeting the target of leanization. For further details please refer to the case study: Impact of the use of exotic compared to local pig breeds on socio-economic development and biodiversity in Vietnam.

Imported pig breeds of Vietnam

Breed Origin Aim of Levels of use Trends utilisation Large White Russia Leanization Normal decreasing Duroc USA Leanization Normal Increasing Landrace USA, Japan, Leanization Wide Increasing France, Russia Yorkshire USA, Japan, Leanization Wide increasing France, Russia Hampshire USA Leanization narrow decreasing Berkshire Russia Leanization Normal decreasing Pietrain Belgium Leanization wide increasing German Noble Germany limited decreasing Cornwall Limited limited Meishan China New imported

Source: Country report Vietnam (2003)

Oceania In Palau all pigs used commercially have been imported - Landrace, Large White and Duroc. Breeds of pigs which were popular in previous years are now in decline namely Hampshire, Berkshire and Poland-China. Locally adapted breeds of pig (the Pacific pig) are very inefficient producers (poor weight gain, low litter sizes) and have been crossed with improved breeds. Pig breeds previously imported to Samoa include the Berkshire, Duroc Hampshire, Large White, Landrace, Poland China and Tamworth. 182 BACKGROUND INFORMATION

A 54: Information from the World Watch List of Domestic Animal Diversity

Proportional share of the world’s total sheep population size and number of breeds in each region

Africa Asia and Europe Latin Near East North The America America Pacific and The Caribbean Population((%) 12.1 38.6 8.2 0.1 0.7 Breeds (%) 11.2 17.7 17.5 6.0 0.4 4.7

Source: Scherf (2000)

Risk status of the world’s domestic sheep breeds recorded

up to 1995 up to 1999 Status Absolute Approx. % Absolute Approx. % Unknown 264 27 391 26 Critical 26 2 62 4 Critical- 8 161 maintained Endangered 59 6 156 10 Endangered- 26 2 43 3 maintained Not at risk 537 55 656 44 Extinct 55 5 181 12 Total 975 100 1495 100

Source: Scherf (2000) REFERENCES BACKGROUND INFORMATION 183

REFERENCES BACKGROUND INFORMATION

Ageeb, A.G. and J.F. Hayes. 2000. Genetic and environmental effects on the productivity of Holstein-Friesian cattle under the climatic conditions of Central Sudan. Tropical Animal Health and Production 32:33-49. Alba, J.D. 1978. Progress in the selection of the Latin American Dairy Criollo. World Animal Review 28:26-30. Allen, J. 2002. Country Report: Australia. ACIAR Proceedings Series Number 108: Development strategies for genetic evaluation for beef production in developing countries. 13-22. Arbeitsgemeinschaft Deutscher Rinderzüchter e.V. 2002. Rinderproduktion, Zucht, Besamung, Leistungsprüfung in der Bundesrepublik Deutschland. Jahresberichte 1972-2002. Arbeitsgemeinschaft Süddeutscher Rinderzucht und Besamungsstationen e.V. 2002. Jahresberichte. 1998-2002. Astruc, J.M., F. Barrillet, M. Fioretti, M. Gabina, E. Gootwine, A.P. Mavrogenis, F.J. Romberg, S.R. Sanna, and E. Stefanake. 2004. Report of the working group on milk recording of sheep. ICAR Report. Results available on the ICAR website www.icar.org. (01.06.2005) Averdunk, G., A. Gottschalk, M. Putz, and E. Rosenberger. 2001. Fleckvieh - Entwicklung zu einer Weltrasse. Arbeitsgemeinschaft Süddeutscher Rinderzucht- und Besamungsorganisationen e.V.München. BLV Verlagsgesellschaft. Blake, R.W. 2004. Dairy cattle performance in difficult environments. 7th World Bruna Conference, Fieragricola, Verona, Italy. 125-131. Blench, A. 1999. Traditional Livestock breeds: Geographical Distribution and Dynamics in Relation to the Ecology of West Africa. London, Overseas Development Institute Portland House. Working Paper 122. Borstlap, R. 1972. Das Fleckvieh als Fleischrind. Der Tierzüchter 24:544-546. Chupin, D. and H. Schuh. 1993. Survey of the present status of the use of artificial insemination in developing countries. World Animal Review 74/75:26-35. Chupin, D. and M. Thibier. 1995. Survey of the present status of the use of artificial insemination in developed countries. World Animal Review 82:58-68. Cossee, O. 2003. Emergency Farm Reconstruction Project Kosovo - Analysis of the Final Survey. FAO, Rome, Italy. Country report of Brazil for FAO’s State of the World’s Animal Genetic Resources. 2003. Ministrerio da Agricultura, Pecuaria e Abastecimento. Country report of Ireland for FAO’s State of the World’s Animal Genetic Resources. 2003. The Department of Agriculture and Food. 184 REFERENCES BACKGROUND INFORMATION

Country report of Japan for FAO’s State of the World’s Animal Genetic Resources. 2003. Editorial Committee Office of the Japanese Country Report, Animal Genetic Resources Laboratory, Genebank, National Institute of Agrobiological Sciences. Country Report of Mexico for FAO’s State of the World’s Animal Genetic Resources. 2003. Secretario De Agricultura, Ganadería, Desarrollo Rural, Pesca Y Alimentación. Country report of the Philippines for FAO’s State of the World’s Animal Genetic Resources. 2003. Country Report of the Russian Federation for FAO’s State of the World’s Animal Genetic Resources. 2003. Ministry of Agriculture of the Russian Federation. Country report of Sri Lanka for FAO’s State of the World’s Animal Genetic Resources. 2003. Department of Animal Production and Health. Country Report of the United States of America for FAO’s State of the World’s Animal Genetic Resources. 2003. USDA, Agricultural Research Service, National Center for Genetic Resources Preservation and National Animal Germplasm Program. Country report of Uruguay for FAO’s State of the World’s Animal Genetic Resources. 2003. Ministerio de Ganaderia, Agricultura y Pesca. Country report of Vietnam for FAO’s State of the World’s Animal Genetic Resources. 2003. Cowert, C. 1983. Brahmans: at home on any range. Foreign agriculture / Foreign Agricultural Service 21:11-12. Dansk Simmental. 2004. Simmental - a World Breed in Denmark. http://www. simmentaldenmark.dk/Engelsk/index.htm. (11.07.2005) Demeke, S., F.W.C. Neser, and S.J. Schoeman. 2003. Early growth performance of Bos taurus x Bos indicus cattle crosses in Ethiopia: estimation of individual crossbreeding effects. Journal of Animal Breeding and Genetics 120:245-257. Ertugrul, O., O. Alpan, M. Umay, A. Bilki, and S. Bulmus. 1999. Improvement of beef production traits of Southern Anatolian Red cattle through crossings with Simmental sires. Veterinarski Arhiv 69:17-28. Eurostat. 2004. EUROSTAT-COMEX Databank. European Union, Luxemburg. Felius, M. 1995. Cattle breeds: an encyclopedia. Doetinchem: Misset. Fésüs, L., L. Sáfár, P. Hajduk, and P. Székely. 2002. The determining role of Merino in Hungarian sheep breeding. 6th Merino World Conference, Budapest, Hungary. Fraga, A. 2004. Personal communication. Giovambattista, G., M.V. Ripoli, P. Peral-Garcia, and J.L. Bouzat. 2001. Indigenous domestic breeds as reservoirs of genetic diversity: the Argentinean Creole cattle. Animal Genetics 32:240-247. Gutierrez, M., B.O. Perez, and M. Milanes. 1985. Estructura y coloración del pelaje de la raza Holstein en el trópico y sus implicaciones productivas. Revista de Salud Animal 7:463-469. REFERENCES BACKGROUND INFORMATION 185

Gyawu, P., K. Asare, and P.K. Karikari. 1988. The performance of imported Holstein Friesian cattle and their progeny in the humid tropics. Bulletin of animal health and production in Africa 36:362-366. Hadi, P.U., N. Ilham, A. Thahar, B. Winarso, D. Vincent, and D. Quirke. 2002. Improving Indonesia’s beef industry. ACIAR Monograph No. 95, vi + 128p. Hansen, H.C. 2004. Dansk Simmental - breeding and management in theory and practice. 15th World Simmental Fleckvieh Congress, 21-30 June 2004, Denmark. Haring, F. 1984. Schafzucht. Ulmer Verlag, Stuttgart, Germany. Holmann, F.J., R.W. Blake, M.V. Hahn, R. Barker, R.A. Milligan, P.A. Oltenacu, and T.L. Stanton. 1990a. Comparative profitability of purebred and crossbred Holstein herds in Venezuela. Journal of Dairy Science 73:2190-2205. Holmann, F.J., R.W. Blake, R.A. Milligan, J S.F. Barker, P.A. Oltenacu, and M.V. Hahn. 1990b. Economic returns from US artificial insemination sires in Holstein herds in Colombia, Mexico, and Venezuela. Journal of Dairy Science 73:2179-2189. Holmann, F.J., R.W. Blake, R.A. Milligan, P.A. Oltenacu, R. Barker, and M.V. Hahn. 1991. Comparative economic returns from artificial insemination and natural service in straightbred and crossbred Holstein herds in Venezuela. Journal of Dairy Science 74:665-677. Hoque, M.A., M.R. Amin, and M.S. Hussen. 1999. Dairy potential of Pabna cows and crossbreds with Sahiwal and Friesian and within- and between-breed sire effects. Asian-Australasian Journal of Animal Sciences 12:161-164. IFAD. 2004. Small Farm Reconstruction and Development Project. http://www.ifad.org/ lrkm/region/pn/ba_1037.htm. (11.07.2005) Ikiror, D. 2001. Heat Stress in Dairy Cattle in Kenya. Tropical Animal Production. Cuvillier Verlag, Göttingen. Islam, S.S. and A.K.F.H. Bhuiyan. 1997. Performance of crossbred Sahiwal cattle at the Pabna milkshed area in Bangladesh. Asian-Australasian Journal of Animal Sciences 10:581-586. Israeli Cattle Breeder’s Association. 2004. Export. http://www.icba-israel.com/icba- export.htm. (11.07.2005) Jabbar, M.A. and M.L. Diedhiou. 2003. Does breed matter to cattle farmers and buyers? Evidence from West Africa. Ecological Economics 45:461-472. Jasiorowski, H.A., M. Stolzman, and Z. Reklewski. 1988. The international Frisian strain comparison trial - A world perspective. FAO, Rome, Italy. Johnson, D.W., A. Tber, and A.T. Benkirane. 1989. Importation au Maroc de bovins Holstein américains: considérations sanitaires. Maghreb Vétérinaire 4:11-14. Künzi, N. and G. Stranzinger. 1993. Allgemeine Tierzucht. 1. Ulmer Verlag, Stuttgart, Germany. Liang, S.I. 1988. Fleckvieh in aller Welt. China. Tierzüchter 40:384. Maleviciute, J., L. Baltrenaite, and I. Miceikiene. 2002. Domestic cattle breed diversity in Lithuania. Veterinarija ir Zootechnika No.20, 87-91. 186 REFERENCES BACKGROUND INFORMATION

Mangurkar, B.R., Y.P. Phadnis, Y.V. Chaudhari, and A.B. Pande. 1986. Lifetime performance of imported Holstein-Friesian and Jersey heifers. Indian Journal of Animal Science 56:1067-1072. McDowell, R.E., G.R. Wiggans, J.K. Camoens, L.D. Van Vleck, and D.G. St.Louis. 1976. Sire comparisons for Holsteins in Mexico versus United States and Canada. Journal Dairy Sciences 59:298. Mudgal, V.D. and C.L. Arora. 1994. Frieswal project: present status and expectations for the future. World Animal Review 79:46-50. Nassuns-Musoke, M. 2002. Environmental Constraints depress the Annual Milk Yield of Frisian Cows on Small Dairy Farms around Kampala. Cuvillier Verlag, Göttingen. Neser, F.W.C., J.B. van Wyk, and G.J. Erasmus. 2002. Simmentaler in Southern Africa - A Scientific Review. http://www.simmentaler.org/Publikasies/SIMMENTALER %20IN%20SOUTHERN%20AFRICA.htm. (11.07.2005) Nsoso, S.J. and T.G. Morake. 1999. A critical look at the use of exotic bulls in traditional beef farming in Botswana. South African Journal of Animal Science 29:100-104. Oklahoma State University. 2004a. Breeds of Livestock - East Anatolian Red. http://www.ansi.okstate.edu/breeds/cattle/eastanatolianred/index.htm. (11.07.2005) Oklahoma State University. 2004b. Breeds of livestock - The Africander. http://www.ansi.okstate.edu/breeds/cattle/africander/index.htm. (11.07.2005) Orbita. 2004. El Simmental en Colombia. http://www.homestead.com/bonavet/files/ page14.html. (11.07.2005) Osorio, A.M. and C.J. Segura. 2002. Reproductive performance of dual-purpose cows in Yucatan, Mexico. Livestock Research for Rural Development 14:1-7. Pangestu, M., Y. Subagyo, P. Yuwonol, B. Rustomo, and G.M. Stone. 2000. Heat tolerance and productivity of local and imported Holstein-Friesian cows in Indonesia. Asian-Australasian Journal of Animal Sciences 13:505-508. Payne, W.J.A. and J. Hodges. 1997. Tropical cattle: origins, breeds, and breeding policies. 1st Edition. Blackwell Science, Oxford. Philipsson, J. 2002. Sustainability of dairy cattle breeding systems utilising artificial insemination in less developed countries - examples of problems and prospects. In: ILRI-SLU 2003. Animal Genetics Training Resource CD-ROM. Version 1. Mwai O. and B. Malmfors (eds.). ILRI (International Livestock Research Institute), Nairobi, Kenya and SLU (Swedish University of Agricultural Sciences), Uppsala, Sweden. Plasse, D., H. Fossi, R. Hoogesteijn, O. Verde, M.C. Rodriguez, R. Rodriguez, and P. Bastidas. 1995. Growth of F1 Bos taurus X Bos indicus beef cattle in Venezuela. II. Initial, final, and carcass weight of bulls, and breeding weight of heifers. Journal of Animal Breeding and Genetics 112:133-145. Rechav, Y., M.W. Kostrzewski, and D.A. Els. 1991. Resistance of indigenous African cattle to the tick Amblyomma hebraeum. Experimental and Applied Acarology 12:229-241. REFERENCES BACKGROUND INFORMATION 187

Robes, B. and A. Martinez. 1990. Development of cattle breeding and milk production in Cuba. Agricultura Tropica et Subtropica 23:29-38. Rocha, A., C. Concha, and D. McKinnon. 1987. Mozambique. Artificial insemination of cattle. Some aspects of breeding policy. World Animal Review 64:12-20. Sadek, R.R., E.A. Helali, M.A. Safwat, S.A.M. Ibrahim, and E.F. Abd. 1994. Evaluation of Friesian cattle performance in commercial farms in Egypt. Egyptian Journal of Animal Production 31:43-64. Schahidi, R., H. Mahmmudzadeh, R. Kasravi, M. Klopcic, K. Malinger, N. Siard, and S. Zgur. 2001. Breeding, artificial insemination and recording of cattle in Iran. 32nd ICAR and INTERBULL Meeting, 16-20 May 2000, Bled, Slovenia. 10:113-117. Scherf, B. 2000. World Watch List for Domestic Animal Diversity. 3rd Edition. FAO publication. http://dad.fao.org/en/refer/library/wwl/wwl3.pdf. (01.06.2005) Schoeman, S.J. 1996. Characterization of beef cattle breeds by virtue of their performances in the National Beef Cattle Performance and Progeny Testing Scheme. South African Journal of Animal Science 26:15-19. Shrestha, J.N.B. 2005. Conserving domestic animal diversity among composite populations. Small Ruminant Research 56: 3-20. Simmental South Africa. 2004. Graphs Simmental South Africa. http://www.simmentaler.org/English/index.htm. (11.07.2005) Sonn, H. 1984. Aspekte der tierischen Erzeugung in Afrika - Erfahrungen beim Einsatz Europäischer Rinderrasen in Afrika. Tropenlandwirt 19:96-105. Sonn, H. 1985. Deutsches Fleckvieh im Ausland - Ein wichtiger Beitrag zur Entwicklung der Welt-Fleckviehzucht. Tierzüchter 37:101-102. Speers, P. 1997. Simmental Boost Income in Sub-Tropical Australian Herds. World Simmental-Fleckvieh Federation. http://www.wsff.info/download-getfile.php?fileID =57. (11.07.2005) Srairi, M.T. and M. El Khattabi. 2001. Évaluation économique et technique de la production laitière intensive en zone semi-aride au Maroc. Cahiers 10:51-55. Stanton, T.L., R.W. Blake, R.L. Quaas, L.D. Van Vleck, and M.L. Carabano. 1991. Genotype by environment interaction for Holstein milk yield in Colombia, Mexico, and Puerto Rico. Journal of Dairy Science 74:1700-1714. Strittmatter K. 2003. Schafzucht. Ulmer Verlag, Stuttgart, Germany. Sutherland, R.A., A.J. Webb, and J.W.B. King. 1985. A survey of world pig breeds and comparisons. Animal Breeding Abstracts 53:1-22. Trail, J.C.M. and K.E. Gregory. 1981. Sahiwal cattle in Africa - how they can help improve milk and beef production. Span. Prog. Agric. 24:28-30. World Bank. 2003. Implementation completion report on a trust fund grant in the amount of $US18.4 Million to the United Nations interim administration mission in Kosovo for the benefit of Kosovo, Serbia and Montenegro (Kosovo) for an emergency farm recontruction project. 26932. http://www-wds.worldbank.org/servlet/WDSContent 188 REFERENCES BACKGROUND INFORMATION

Server/WDSP/IB/2004/01/08/000090341_20040108095652/Rendered/PDF/269320 YF.pdf. (11.07.2005) 189

GENE FLOW IN ANIMAL GENETIC RESOURCES. A STUDY ON STATUS, IMPACT AND TRENDS

Editors: Anne Valle Zárate, Katinka Musavaya and Cornelia Schäfer

Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, Germany

ANNEXES 9.2-9.5

190

ANNEX 191

ANNEX

9.2 Global gene flow of sheep

C. Schäfer, A. Valle Zárate 192 TABLE OF CONTENTS

TABLE OF CONTENTS

List of tables 193

List of figures 193

Abbreviations 194

1 Introduction 195

2 Historical development of sheep gene flow 195 2.1 Gene flow during domestication and breed formation 195 2.2 Influence of human migration on transfer of animals 198 2.2.1 Europe 199 2.2.2 Africa 200 2.2.3 Asia 201 2.2.4 The Americas 201 2.2.5 Oceania 202 2.3 Influence of breeding in further development and spread of breeds 203 2.3.1 Influence of breeding on development and spread of the Merino breed 203 2.3.2 Influence of breeding in development and spread of other temperate breeds 207 2.3.3 Influence of breeding in development and spread of breeds in developing countries 207 2.4 Influence of technology and global mobility on the dissemination of genetic material 209

3 Current status and actual trends of sheep gene flow 210 3.1 Gene flow indicated by selected export and import 210 3.1.1 EU-15 countries 210 3.1.2 Eastern Europe 212 3.1.3 America 213 3.1.4 Oceania 213 3.1.5 Asia 213 3.1.6 Middle East 214 3.1.7 Africa 214 3.2 Gene flow indicated by the introduction of foreign genetic material 214 LIST OF TABLES 193

3.2.1 Europe 215 3.2.2 America 218 3.2.3 Oceania 220 3.2.4 Asia 220 3.2.5 Middle East 220 3.2.6 Africa 221 3.2.7 The Booroola gene - An example for the introduction of foreign genetic material 221 3.3 Gene flow indicated by Country Report excerpts 223

4 References global gene flow of sheep 225

LIST OF TABLES

Table 1: Systematic classification of the genus Ovis 195 Table 2: Characteristics of the Merino breed and tropical hair sheep breeds according to production environment, purpose and regions of dissemination 196 Table 3: Fine wool breeds of Russia and their formation basis 206 Table 4: Distribution of sheep population in the EU-15 countries including export and import of breeding animals in 2003 211 Table 5: Russian sheep export and import figures of sheep 213 Table 6: Breeds used in Merino breeding in the past and the recent Merino breeds in Hungary 215 Table 7: Changes in breed structure in sheep population of Poland between 1976 and 2000 217 Table 8: Source and year of importation of Booroola gene into European countries 222 Table 9: Country Report excerpts on gene flow of sheep 224

LIST OF FIGURES

Figure 1: Development of fine wool sheep with uniform fleece composition in comparison to double coated hair sheep breeds, from domestication to dissemination of Iberian Merino 198 Figure 2: Spread of the Merino breed after fall of the Spanish Merino monopoly 200 Figure 3: Development of exports and imports of breeding sheep in EU-15 countries from 1992-2003 212 Figure 4: Geographic illustration of the flow of the Booroola gene 223 194 ABBREVIATIONS

ABBREVIATIONS

BC before Christ BMPIB Bone morphogenetic protein IB receptor CSIRO Commonwealth Scientific and Industrial Research Organisation DAD-IS Domestic Animal Diversity Information System EU-15 Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Portugal, Spain, Sweden, The Netherlands, United Kingdom FAO Food and Agriculture Organisation of the United Nations FRG former Federal Republic of Germany GDR former German Democratic Republic ICAR International Committee for Animal Recording PROFOR Program on Forests (multi-donor trust fund program housed at the World Bank) UK United Kingdom USA United States of America USSR former Union of Socialist Soviet Republics

INTRODUCTION 195

1 INTRODUCTION

For this chapter Merino and Hair Sheep breeds have been chosen to highlight typical developments in sheep. The Merino is a wool sheep which used to be important worldwide in both crossbreeding and purebreeding programmes. Hair sheep have adapted to hot climates. The influences of domestication, breed formation, human migration, and breeding methods on the diversification of sheep were studied. Reference to other sheep breeds are made where appropriate. Data was sourced from the Eurostat statistic database, which, however, does not contain any breed information, and Country Reports on the State of Animal Genetic Resources. Additionally project reports and publications were studied and breeding organisations and experts from different countries were contacted.

2 HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW

2.1 Gene flow during domestication and breed formation The world sheep population today accounts for 1,024 million sheep. 350 million, about one third, are found in developed countries, the remaining two thirds in developing countries (FAO, 2004). The systematic classification of the genus Ovis is still an open question. It is generally accepted that it can be subdivided into one domestic species (O. aries) and several wild species with some of them differing in chromosome number (Table 1). Of the wild species the Asiatic mouflon (O. orientalis), European mouflon (O. musimon) and the Urial (O. vignei) are believed to be ancestors of domestic sheep.

Table 1: Systematic classification of the genus Ovis

Common name Chromosome number Systematic name Asiatic mouflon 54 O. orientalis European mouflon 54 O. musimon Urial sheep 58 O. vignei Arkhar argali sheep 56 O. ammon Snow sheep 52 O. nivicola Dall sheep 54 O. dalli Bighorn sheep 54 O. canadansis Domestic sheep 54 O. aries (about 1000 breeds) However, there are also suggestions that domesticated sheep all belong to the species Ovis ammon L. with 28 Eurasian and 9 North American subspecies. This theory was supported by the fact that all wild and domesticated sheep are interfertile (Strittmatter, 2003). The Asian and European Mouflon were found in Europe and West Asia and are known as the ancestors of many sheep breeds. The Urial sheep were found in the north up to the Caspian Sea, in the east to Cashmere and in the south to the Oman. Characteristics of the 196 HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW wild ancestors that can be traced to modern breeds and that allow the differentiation of later breeds are horns in both sexes, and a short tail. In comparison to the Mouflon, the Urial are taller and longer with almost round horns in adult rams and have a tail with 8-10 more vertebrae (Strittmatter, 2003). Sheep are believed to have been domesticated about 9000 - 7500 BC with the original centre of domestication in the Aralo-Caspian steppe (Zeuner, 1963). A second centre probably came up around 7000 - 6000 BC in Greece and East Europe (Ludwig, 1997). However, the origin is not fully clear. The first breeding steps after domestication resulted in sheep with a high genetic variability seen particularly in coat colour, but also in horn shape, fleece architecture and other visible traits. With at least 1,747 known sheep breeds in the world, domesticated sheep, compared to their wild ancestors, are characterised by an impressive diversity (DAD-IS, 2004). This diversification has been driven by factors affecting breed formation, the occurrence of the characteristic wool fleece, the occurrence of the fat tail and in particular the adaptation to a very broad range of environmental conditions when sheep spread all over the world (Haring, 1984). Among all domestic animals, sheep show the strongest relation to their production environment which is reflected by the great diversity of sheep breeds. Their ancestry, their local environment and the purpose of their keeping led to the diversification - for example the development of fat-tail or milk sheep. Sheep are generally bred for wool, meat, fat, fur, skin, and milk and are in general at least dual-purpose breeds. Table 2 below summarises the classification of the two breeds respectively breed groups at focus, the fine wool sheep and tropical hair sheep, according to production environment and purpose. Both stem from long tailed Urial sheep.

Table 2: Characteristics of the Merino breed and tropical hair sheep breeds according to production environment, purpose and regions of dissemination

Breed group Production Production purpose Main regions environment Tropical hair sheep Hot-humid Meat production, skin West Africa, Southeast environments with as side-product Asia, The Americas severe production constraints Fine wool sheep Continental pastures Wool production, Australia, South with uniform additionally meat Africa, South distribution of rainfall America, continental and constant fodder Europe supply Sheep were domesticated just as their wild ancestors developed a double coat, with a short hair coat of kemp fibres and frequently grown and shed underwool fibres. Domesticated sheep breeds with this original double coat are referred to as “hair sheep”. The first “primitive wool sheep” with increasingly uniform coat composition go back to the 6th millennium BC in the East Iran. These developed from the original double coat sheep with primary follicles producing kemp fibres and secondary follicles producing extremely fine underwool and exhibiting seasonal growth and shedding processes, to a uniform wool sheep with continuously growing wool fleece uniformly produced from primary and secondary follicles. This process seems to have been going on for some thousands of years HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW 197 during which the primary follicle changed from producing kemp fibres to wool fibres, and the seasonal growth and shedding of the undercoat changed to a continuous growth requiring shearing. This lead to increasingly uniform fibres within the coat and increased the amount of wool harvested. The consequence of this ongoing process is the known great variety of wool sheep breeds with differing fleece composition and characteristics. Such mutations must have occurred in descendants of both the Mouflon and the Urial sheep, which explains the existence of short and long tailed wool sheep with different fleece characteristics (Epstein, 1971; Haring, 1984; Ryder, 1984; Schäfer, 1998; Strittmatter, 2003). Henseler (1944) supposed that the occurrence of uniform-fine-wool sheep depended on further mutations of coarse woollen descendants of the Urial. When man discovered that wool fleeces could be used to produce fabrics of high quality, sheep in the Middle East lost genetic originality. This resulted in marginalised traditional populations of sheep at least partially standardised woolly and white sheep (Lauvergne, 1990). The preferential dissemination of fine woollen sheep increased fineness, but also replaced local stock with fine wool sheep where the environment was suitable for fine wool sheep. The fine wool sheep of early extreme quality that was bred in Phrygian in the 7th and 8th century BC to produce the Milet-Tuch was carefully prevented from spreading to maintain a monopoly. How the first breeding animals, later referred to as “Merino”, occurred in the Iberian Peninsula is not known. However, due to trading connections between the Orient and Spain, it became a centre of sheep breeding in the following centuries. From Spain, these Merinos began their triumphal distribution to France, Germany, Australia, Africa, South America, and the rest of the world (Ryder, 1984; Haring, 1984). Today the Merino breed includes fine, medium, and strong wool strains. However, wool-bearing sheep were initially only of interest in climates where the additional value of wool was seen in the insulation purpose, mainly to make human clothing. On the other hand, in hot and humid climates large amounts of wool is a major disadvantage for sheep. Therefore, in tropical environments wool breeds did not successfully replace breeds with hair coats. This hair coat is often regarded as primitive due to its early appearance in domestication, but really it can be regarded as the result of thousands of years of adaptation of these hair sheep breeds to their extreme production environments in which other sheep breeds do not often survive (Schäfer, 1998). Figure 1 summarises the formation of domestic sheep breeds with the two extreme coat characteristics, the double coated hair sheep on one hand and the uniform fine wool sheep on the other. 198 HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW

Figure 1: Development of fine wool sheep with uniform fleece composition in comparison to double coated hair sheep breeds, from domestication to dissemination of Iberian Merino

Mouflon and Urial development partial replacement in domestication suitable environments 9000-7500 BC in Arab-Caspian Steppe 7000-6000 BC in Greece and East Europe

mutation hair sheep

“primitive wool sheep” hair sheep breeds (development towards uniform wool fleece started) 6th millennium in East Iran

preferential dissemination of wooly and white sheep

“fine wool sheep” (uniformity of wool fleece wool sheep with achieved) mixed fleece 3000 BC in Sumerian and composition Phoenician

preferential dissemination of sheep with uniform wool fleece

“extreme fine wool sheep” (increased fineness of wool) 7th and 8th century BC in Phrygia

prevention of free dissemination

Iberian Merino sheep with mixed hair sheep fleece composition

2.2 Influence of human migration on transfer of animals The biggest catalyst of gene flow after domestication were tribal migrations during the Neolithic transition, taking along their domesticated animals. At that time, sheep were generally hair sheep. Mutations towards uniform wool fleeces took place only after the first migrations and can be regarded as regional developments. Sheep with favourable characteristics were traded and replaced breeds in other regions. The migration of people and livestock has not generally been in one continuous direction. In most regions there has been fairly constant trade in livestock from one community to the HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW 199 next throughout human history, modifying every population through a gradual inflow of genes (Henson, 1992). Therefore, the influence of dissemination based on migration compared to that based on trading favourable breeding animals cannot always be clearly distinguished. Through human migrations during the Neolithic period, sheep spread from the centre of domestication into what is now Iran, eastwards into the Indian subcontinent and Southeast Asia, westwards into western Asia and on into Europe and Africa.

2.2.1 Europe

In Europe the original hair sheep were eventually replaced by wool-bearing sheep once they had developed, due to their great importance in producing insulating fabrics. How the first “fine wool sheep” occurred in Europe, namely in Greece, Italy and Spain is not known. Some essayists suggest the breed’s ancestry was the Caspian area, and that they came with human migrations to the Iberian Peninsula across the Mediterranean Sea. It may be that the Iberians brought them with them on their way from the Caucasus to Spain. It is accepted that the “fine wool breed”, later referred to as “Merino”, was created in the southern Iberian Peninsula, precisely in the area in which nowadays the main nucleus is located (Baraja, 2002). Here, the Merino breed stayed isolated from the rest of the world until the end of the 18th century and exclusively exploited the potential of the Merino breed for producing fine wool by forbidding Merino sheep exportation (Baraja, 2002). After the Arabs brought Islam to Spain, the trading connections to the Orient made Spain a centre of sheep breeding. Countries across the whole world became interested in the development of this “fine” wool breed. The enormous demand for Merino breeding animals was satisfied in the last quarter of the 18th century and the beginning of the 19th century. The breed spread along different countries into western and eastern Europe. Where it found a continental environment suitable for its management and development, it began its triumphal distribution (Haring, 1984). In western Europe, mainly France, Germany, Austria, and Italy, important Merino flocks were developed. In eastern Europe, the Merino breed gained importance in Russia, Poland, the Czech Republic and Hungary. Countries with more maritime climates, such as Holland, Great Britain and Scandinavia, proved unsuitable for the breed. Figure 2 summarises the most important aspects of the spread of the Merino breed after the fall of the Spanish monopoly. More detailed information on national developments of the breed can be found in Annex 9.1 A 2.

200 HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW

Figure 2: Spread of the Merino breed after fall of the Spanish Merino monopoly

Europe United Kingdom 1897 Netherlands Hungary 1786 Germany 1774 1867 1765 France Spain 1786 North America Austria 1784 Canada/USA 1850

1863 1864

Argentina 1797 Africa 1786 Australia 1869 Uruguay South Africa Oceania South 1873 New Zealand America 1773

Source: own elaboration, data from Haring (1984)

Where Merino wool sheep were imported, the breed was usually established nationally by crossbreeding it with local indigenous breeds. The sustainable establishment was often based on follow-up imports of purebred stock. Additionally, purebred nucleus flocks were started from which dissemination took place. Germany, Austria, and France - as well as Spain - founded Merino breeds which later gained important genetic influence on breeds around the world. One outstanding example is the Rambouillet breed in France, created by crossbreeding Spanish Merino rams with local indigenous ewes (Annex 9.1 A 2). With the increasing importance of mutton production, beside the Rambouillet-Merino, the dual- purpose Merino Precoce was developed with two types, the Merino Chatillon and Soissons. Both types influenced the development of the German Mutton Merino sustainably whose influence went beyond regional importance (Haring, 1984).

2.2.2 Africa

A great variety of sheep breeds exist in Africa, developed by differentiation from the hair sheep breeds indigenous to Africa since Neolithic times, and also through the later arrival of mostly wool breeds. Remains of domestic sheep (and goats) have been found dating 4800 BC (Payne and Hodges, 1997). Around the Sahara, the African Long-legged type predominates, which were probably introduced by Hamitic pastoralists from western Asia. The two separate African groups of thin-tailed hair sheep (the larger Savannah type and the smaller Forest type) may have developed from the same source by natural and man-assisted selection in their two widely different environments (Payne and Wilson, 1999). The fat-tailed coarse wool sheep appear to have been introduced into Africa some 3,000 years after the thin-tailed hairy breeds, probably via the Isthmus of Suez and the Straits of Bab-el-Mandeb at the northern and southern end of the Red Sea. These fat-tail coarse HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW 201 woollen sheep replaced thin-tailed hairy sheep in Egypt and in what is now Libya and Tunisia. Those entering what is now Somalia and East Africa were eventually taken by their migrating owners to the far south of Africa, and their descendants are the present fat- tailed breeds of Northeast, East, Central and South Africa. The origin of the fat-rumped breeds remains unknown but they occur in two separated and ecologically different regions, which are, on the one hand Northeast Africa, East Africa, the Indian Ocean and Red Sea littorals of the ; and on the other hand Central Asia. In South Africa, breeds for the harsher areas developed from the fat-rumped sheep of Somali-Arabian origin, which were introduced into the Cape Province of South Africa during the 19th century and were quickly improved to form the breed known as the Blackhead Persian (Payne and Wilson, 1999). From here it has also spread to other parts of South Africa and farther north, notably to Tanzania, Kenya, Ethiopia, and even Ghana (Gatenby, 1986). It has been used for crossbreeding, and the best known result is the Dorper synthetic breed. The Dorper breed was developed in the Grootfontein area of South Africa from 1942 onwards by crossing Dorset Horn males with Blackhead Persian females. A fixed type was developed through inter-se mating. A breed society was established in 1950 in the Republic of South Africa. The Dorsian, a white variety, was affiliated to the Dorper in 1964. Some specialisation for different functions and markets is taking place. The Blackhead Persian is used in a variety of other crosses, which have characteristics similar to the Dorper (Wilson, 1991). But Blackhead Persians have also been introduced to Central and South America for crossbreeding purposes (Wilson, 1991).

2.2.3 Asia

In Asia, a variety of sheep breeds exist which are derived from the hairy breeds that arrived in Asia with migratory tribes. The influence of wool breeds must come from later migratory events. Fat-tailed breeds with unknown coat characteristics thrive in the drier regions of northern India. Farther south fat-tailed and thin-tailed wool types were found which are gradually being replaced by thin-tailed hairy breeds. The humid regions of Southeast Asia have few sheep breeds but most are small hairy types. Exceptions are a fat-tailed type of Western Asia, possibly from India, and a poor wool sheep breed from crossing the indigenous hair type with wool sheep from Australia or South Africa. The poor coarse wool sheep of southern Thailand and Malaysia appear to have originated from crosses between woollen Chinese sheep, taken south by Thai and other peoples, and the native hairy sheep of the region (Payne and Wilson, 1999). In China alone, a great variety of breeds is found. The authors do not know when the Merino breed arrived in China but it has became the most numerous and the nation now has the world’s largest sheep population.

2.2.4 The Americas

There are no domesticated sheep truly indigenous to the Americas. There have been two migratory events that have brought domestic sheep to Central America and the Caribbean. Firstly, the sheep type generally known today as Criollo type arrived with early settlers from Spain and Portugal in the 16th century. These were wool sheep, but under tropical conditions much of the fleece has been lost or reduced. 202 HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW

A second source was hair sheep from Africa. These gave rise to most of the truly woolless sheep of the area, established in the humid tropical areas where the Iberian sheep had never become acclimatised. The hair sheep in the Americas that are of African origin came from countries between Nigeria and Angola in West Africa during the slave trade during the 17th and 18th century. It is supposed that two types of thin-tailed sheep came to the Americas. One, without mane, probably came from savannah zones. The other type of thin-tailed sheep, which gave origin to the hair sheep carrying a mane and throat ruff, probably came from the Djallonke of Cameroon (Shelton and Figueiredo, 1990; Fitzhugh and Bradford, 1983). Hair sheep populations are found on many Caribbean islands, in Central and South American countries around the Caribbean basin, and in north-eastern Brazil. Both Barbados and the Virgin Islands have exported hair sheep to the United States (Fitzhugh and Bradford, 1983). Studies on the genetic relationship of hair sheep breeds of the Americas by Muigai et al. (2002) however suggested genetic influence of different origins. The study indicated that the major genetic influence in three American hair sheep populations (Barbados Blackbelly, St. Croix, and Gulf Coast Native) is likely European and may have originated from crossbreeding with Merino breeds. As the chronologically youngest source, most of the traditional breeds of wool sheep of European origin have been introduced to tropical Americas. In a few cases, these are used as a basis of commercial programmes of limited scope, but for the most part these sheep have not persisted and are of no importance in tropical regions. But they may have contributed genes to some of the present strains (Shelton and Figueiredo, 1990) as supported by the findings of Muigai et al. (2002). Merinos were introduced to South America by the early 19th century. In countries with cold and dry highlands, such as Argentina, Uruguay, Paraguay, Brazil and Chile, the Merino quickly gained importance. In Argentina for example, the Merino breed became the dominant breed (Shelton and Figueiredo, 1990).

2.2.5 Oceania

Domestic sheep in Australia arrived with the First Fleet in 1788 and were fat-tailed sheep. However, the sheep industry in Australia is based largely on the Australian Merino, which goes back to the Spanish Merino arriving in 1797 from the Cape Colony. Initially, the expansion was along the green pasture belt along the coastline. However, when the wool industry - due to its outstanding wool quality - grew, the traditional pasture was used up. The most significant development occurred when these pioneering breeders broke away from the coastal line and moved their flocks into the drier, harsher inland country that under normal circumstances would not have been considered suitable for the fine wool that the industry aimed to produce (Padbury, 2002). The first sheep in New Zealand were landed by Captain Cook in 1773, also of the Merino breed. However, only when the gold fever from the 1860s to the 1880s drew thousands of prospectors to New Zealand did large sheep farms began to be established on land cleared from the native forests. Many Australian squatters came over to lease large areas for their flocks. Thousands of sheep, predominantly Merino, were shipped from Australia. This period is known as the wool period. However, sheep scarp came with the Australian flocks in the 1860s, killing thousands of sheep. The Merino breed today makes up only a small portion of the New Zealand breeds. However, it is the only breed to thrive on the high altitude pastures of native grass (New Zealand Merino Breeders Association, 2005). HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW 203

2.3 Influence of breeding in further development and spread of breeds Three apparent steps of breeding can be distinguished: spontaneous mating, planned mating and planned mating plus selection or crossbreeding programs. Breeding aims at developing individuals and populations with improved characteristics. Whenever an improvement is achieved which goes beyond national importance, gene flow may start. As certain breeding approaches are younger than others, the examples given below partially overlap with the information assembled in chapter 3.2 on indicators of current status of gene flow, and repetitions occur.

2.3.1 Influence of breeding on development and spread of the Merino breed

The Merino breed provides an excellent example of how breeding influences the development and spread of a breed with international importance can be exemplary depicted for the Merino breed. When the Merino was first exported from Spain at the end of the 18th century, they were generally crossbred with local breeds in the destination country to improve wool quantity and quality. Where the breed found an environment suitable for its management and development, distinct breeds developed, adapted to the specific local conditions. Due to the exceptional importance of fine wool production at that time, Merino’s were generally bred for wool production for garments and the breeding goals were to improve quantity and quality. In many countries, the Merino breeds kept their landrace character as they were kept under most extensive management in regions with extremely specific conditions. It was therefore important to combine the specific adaptation of the local landrace with the improved production traits of the Merino. This way, the successful distribution of the Merino breed was possible, being able to equally produce on continental pastures with uniform distribution of rainfall and constant fodder supply, on drier and harsher inland countries such as Australia and South Africa, and on the high altitudes of some South American countries. Soon after World War II, fine wool sheep breeding and activities related to managing Merinos came under severe economic pressure when the use of wool fibre substitutes in fabrics and the development of artificial fibres influenced the world wool market in such a negative way that the wool prices dropped drastically. The only solution was a quick orientation towards meat production world wide. Wool value, which still makes up the largest part of the profits from Merino sheep productions, has been falling ever since. Again, there were two ways to improve meat production from Merino sheep, selection or crossbreeding. France for example had already seen in 1860 a wool price reduction of two thirds and a meat price increase of one third (Haring, 1984). It pioneered the breeding goal of adapting their fine wool breed to an intensive dual-purpose breed, and developed an early maturing Merino Precoce with two types, the Merino Chatillonais and the Merino Soissonnais. Two decades later, Germany developed their mutton type by crossbreeding with these two meat- types. The German Mutton Merino were very successful and as a result it was recognised that with this breed, fine wool could indeed be produced by meat sheep (Haring, 1984). In Spain’s Merino production, for example, today’s profit from wool has reached a low of about 2% of the whole production (Baraja, 2002). 204 HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW

In Hungary, Merino sheep have been raised since the middle of the 1700s. The predecessor of the modern Hungarian Merino sheep, the Combing Merino, emerged from crossbreeding the Electoral-Negretti Merino with the local breeds by the 1920-30s. This cross was subsequently upgraded using Rambouillet and German mutton breeds. The heterogeneity of Hungarian Combing Merino was further increased by upgrading with Soviet Merino breeds. These included sheep from the Caucasus, Stavropol, Groznyy, and most of all Askania. From the 1960s onwards meat purpose breeding also got underway, first with Merino Precoce from France and with German Mutton Merino, both from East and West Germany. During the 1970-80s the Hungarian Combing Merino was further improved by crossing with the long-wool Kent and Corriedale breeds, with Australian Merinos to improve wool quality and with Booroola Merino to increase reproduction rates. In order to standardise the Hungarian Merino Breed, a new flock book for the Hungarian Merino was established in 1993. Meanwhile, sub-breeds were excluded from breeding. And following the introduction of the Animal Breeding Law, offspring from crossings with another breed can be re-entered into the main part of the flock book after four generations (Fésüs et al., 2002). In China, the Merino breed arrived a long time ago and was crossed with local breeds. The most numerous cross is the Xianjiang Merino, a composite of Caucasian, Precoce, Hazake, and Mongolian, which is known for its good quality meat and wool. The northeast Chinese Merino has been bred more towards meat as a composite of Rambouillet, Corriedale, and Mongolian with top quality fleece and meat. The Mongolian or Mongol Merino or China fat-tailed is a composite of the Soviet Merino and the Caucasian. Due to its fat-tail it is well adapted to harsh environments with sudden changes. The north-western Chinese Merino is a composite of Caucasian and Xinjiang Merino x Mongolia Tibetan, which, also known as Gansu Alpine Fine Wool sheep, is characterised by its exceptional fine wool (DAD-IS, 2004). In the 1950s, China imported about 20,000 German Mutton Merino from the former East Germany. In the 1990s further imports followed (Schmidt, 1995). From these imports it can be concluded that China’s Merino breeding goals are switching from wool to meat production. The Australian Merino provides a unique example of top level sheep breeding with an industrial structure, which allows the implementation of intensive selection schemes. The advantage for genetic improvement is achieved in elite ram breeding nuclei of an open nucleus system - also referred to as centres of genetic improvement - and transferred to commercial flocks. An elite flock of 2,000 breeding ewes can provide sufficient rams to commercial flocks totalling about 100,000 breeding ewes. In other words, the effect of selection in the elite flock is magnified by a factor of 50 in the industry. The logic behind this structure lies in the concentration of superior animals (and thus genes) in the upper tier of the hierarchy. The genes in the upper tier are multiplied and reach the base of commercial flocks via the transfer of rams (Ponzoni, 1992). In Australia, years of breeding work led to the evolution of different strains suitable for varying climates and pastures. The most important one is the Peppin Merino for which it seems clear that Merinos of French and Spanish origin were introduced. 70% of today’s Australian Merino are said to be directly descended from the Peppin. It is known for its large frame and its heavy fleece in the midrange of Merino wool qualities and was developed for the temperate climates of the country. The South Australian Merino were specifically bred to provide an economic return from wool under arid pastoral conditions. HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW 205

It is physically the largest of the Merino strains and less wrinkled than other strains of the country. The fleece is at the strongest end of the range of Merino wool types. The Saxon Merino are found exclusively in the higher rainfall areas of Australia and physically contrasts with the South Australian Merino. It is the smallest Merino type, lowest wool weights but it is without peer in the quality of wool. The wool is extremely white and bright, soft and fine. Additionally, there is a Fine Wool type. It is distinguished by a small to medium-sized but compact frame and produces a soft, bright coloured dense fleece (Handbook of Australian Livestock, Australian Meat and Livestock Corporation, 1989). Early last century, larger plainer bodied sheep began to replace the wrinkled type that had been popular previously. Two centuries of skilled breeding has developed the Merino amazingly. It now produces far more meat and much better wool than its forebears. In 1800, the average wool cut of a pure blood Spanish Merino was less than 2 kg of wool. Today, most of the ewes in the Merino studs in Australia produce more than 7 kg. At the same time, the Australian Merino has adapted to dozens of different environments; from the high rainfall pastures of Tasmania to the drier harsher almost semi-desert areas of inland Australia (Padbury, 2002). The former USSR is a striking example where numerous sheep breeds have been formed by crossing imported and national sheep breeds. Crossbreeding programmes have been started to form new breeds combining the genotypes of high producing breeds and the characteristics of the local populations such as strong constitution, hardiness and adaptation to mountainous environments. For example, the Merino breed has influenced Russia’s fine wool sheep. Table 3 summarises the different fine wool breeds and their genetic basis of formation. 206 HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW

Table 3: Fine wool breeds of Russia and their formation basis

Fine wool breeds Formation basis 1. Wool breeds Grozny Purebred Australian Merino, in addition Australian Merino rams x Novocaucasian and Mazaev Merino ewes (local Merinos) Salsk American Rambouillet rams x local Merino ewes Stavropol Interse mating of the local Merinos, additionally crossing with American Rambouillet 2. Wool-mutton type Altai (earlier Siberian Based on local Merinos brought to Siberia from the North Caucasus, Rambouillet) mated to American Rambouillet rams, later crossing with Australian Merino Askanian Local Merino ewes mated to American Rambouillet rams, selection and infusion of small quantities of Precoce blood Caucasian Improved local Merino ewes crossed with American Rambouillet and Askanian rams Kirgiz Finewool Coarse-wooled fat-rumped ewes crossed with finewool rams, first of the local Novocaucasian and Siberian Merino breed and then of the Precoce and the Wurttemberg (Merino Landschaf) from Germany; following inter se mating and mating to rams of the Caucasian, Altai, Stavropol, Askanian and particularly Grozny type Krasnoyarsk Finewool Based on local Merinos from the North Caucasus and American Rambouillet ewes and rams; further crossbreeding with Precoce from Germany; later crossbreeding with Askanian and Grozny rams North Kazakh Merino Crossbreeding coarse-wooled fat-ramped ewes with Novocaucasian Merino, Rambouillet and Precoce, subsequent crossing with Altai, Askanian and Grozny rams South Kazakh Merino Crossbreeding fat-rumped ewes with Novocaucasian and Soviet Merino rams, then to Caucasian, Grozny, and Stavropol breeds, additionally crossbreeding with small numbers of Altai and Askanian Soviet Merino (varieties Crossbreeding local coarse-wool sheep from various parts of the country North Caucasian and with finewool rams of different breeds: initially American Rambouillet Siberian) and Australian Merino, later Caucasian, Stavropol, Grozny, and Altai Trans-Baikal Finewool Based on Mongolian and Buryat coarse-wooled fat-tailed sheep unsuccessfully crossbred with Electoral and Infantado sheep, later Precoce, Novocaucasian and Siberian Merinos were brought in, subsequent crossbreeding with Precoce, Altai, and Grozny rams 3. Mutton-wool type Georgian Fat-tailed Crossbreeding local Tushin ewes with Soviet Merino rams; subsequent Finewool crossbreeding with Caucasian rams Kazakh Arkhar-Merino Inseminating Novocaucasian ewes with Arkhar rams; crossbred rams were crossbred with Precoce and Rambouillet ewes; Kazakh Finewool Crossbreeding selected fine-wooled fat-rumped ewes with American Rambouillet rams Volgograd Crossbreeding fat-rumped ewes to Precoce rams; further crossbreeding with Caucasian and Grozny rams Source: adapted from Dmitriev and Ernst (1989) HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW 207

2.3.2 Influence of breeding in development and spread of other temperate breeds

Apart from the focus of this global study on the Merino breed, other breeds developed which are of major regional and global importance. In particular British breeds, developed through the early efforts of the British breed societies, have influenced the whole New World. Suffolk, Lincoln, Romney Marsh, Corriedale, Poll Dorset, Border Leicester, Scottish Blackface, and many others are reported to have been transferred to many other countries across the world and have impacted national population compositions in different ways. The Ile de France, Charolais, and Texel and breeds from Russia are other European breeds whose historic and current global spread is evident. Scherf (2000) reports in The World Watch List published by the FAO based on figures from 1999, that Europe held 17.5% of the world’s sheep population but 47.9% of the world’s known sheep breeds. These figures reflect the effective breeding work in Europe and the subsequent influence of these breeds globally. Greece, Italy, Spain and France have large populations of milk sheep. An ICAR report on Dairy Sheep 2004 (Astruc et al. 2004) documents intensive purebreeding efforts on developing milk sheep breeds with countries having above 30% of recorded females under official milk recording. Looking at breed level, the East Frisian Milk sheep from Germany, the Lacaune from France, and the Awassi respectively Assaf breed from Israel have spread beyond national importance.

2.3.3 Influence of breeding in development and spread of breeds in developing countries

In general, in developing countries spontaneous mating is widely spread among sheep. Breeding work faces many challenges to introducing planned mating and selection or crossbreeding programmes. Subjective selection criteria often replace objectively obtained and economically assessed selection criteria. Examples of selection programs based on objectively measurable and repeatable traits are generally rare in the tropical and subtropical regions. The Karakul breeding program in Namibia was one example where selection based on performance testing of an objectively measurable lamb coat trait was very successful (Horst and Reh, 1999) and generated international demand for the breeding material. Improving milk production through selection based on performance testing of objectively measurable repeatable traits was another example. However, due to the high level of organisation required, such performance testing is only found in countries with advanced breeding organisation such as Israel and some South European countries. Introducing improved European meat breeds was initially regarded as the key to improving meat production in developing regions of the tropics and subtropics. Examples were numerous but usually barely successful (Horst and Reh, 1999). In general, the lack of adaptability to the constraints of local production systems, namely to high ambient temperatures, restricted availability of water and fodder, and susceptibility to local diseases does not allow the exotic breeds to make the most of their productive potential and outperform the indigenous breeds. For example, Horst and Reh (1999) compiled information on the reproductive performance of exotic breeds in subtropical production 208 HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW environments, and they state that adapting the breeding season of exotic breeds was a major problem. Therefore, establishing purebred flocks of these imported breeds in the tropics and subtropics was rare and restricted to the edges of these regions. However, these purebred flocks are essential to produce F1-ewes or lamb end products. Alternatively, combination crossing can be used where a breeding scheme can be successfully carried out. The most important example is the development of the synthetic Dorper breed which combines the adaptability of the Blackhead Persian to the hot environment on the ewe side, and the growth performance of the Dorset breed on the sire side. Horst and Reh (1999) stated that the Dorper breed was the most important meat breed in the extensive production systems of South Africa, Botswana, Zimbabwe and Kenya. Some specialisation for different functions and markets has taken place. A detailed description of the history of the Dorper sheep was presented by Milne (2000) and de Waal and Combrinck (2000). The Blackhead Persian is used in a variety of other crosses, which have characteristics similar to the Dorper (Wilson, 1991). In producing a new breed, the initial crossings must be on a large enough scale (using at least ten rams) to avoid the deleterious effects of inbreeding. The characteristic of this breeding strategy is that gene flow takes place several generations before the new breed has been established (Gatenby, 1986). An example for the successful three-way crossing including an exotic breed from a temperate climate in a subtropical environment is the Assaf breed in Israel, where East Frisian Milksheep were crossed with Improved Awassi sheep. Details are given in the case study. Information on the global formation of composite breeds is compiled by Shrestha (2005). He stated that in the past centuries, more than 443 composite breed populations of sheep have been developed in 68 countries of the world. Details of many composite breeds world wide including country of formation, composite breed name, number of foundation breeds, and year of origin were given (Annex 9.1 A 4). According to the author, they all derived from two or more distinct populations. The breeds chosen world wide for potential genetic merit were the Dorset, Suffolk and Texel for meat, the Finnish Landrace, Romanov and Booroola for fecundity, the Merino and Rambouillet for wool, the Karakul for fur, and the Awassi and East Frisian for milk. For North America, Shrestha (2005) illustrated the influences of different breeds at different times in developing composite breeds. In the 1960s, the Finnish Landrace had been identified as having considerable potential to improve reproduction of the sheep breeds established in the USA. The Polypay breed was developed by reciprocal crossing between the Dorset x Targhee breed and the Finnish Landrace x Rambouillet in the 1970s and inter se mating of the four breed cross. Canada imported the East Friesian, Finnish Landrace and Ile de France breeds from continental Europe and assembled these breeds with the Corriedale, Dorset, Leicester, Lincoln, North Country Cheviot, Romnelet, Shropshire, Southdown and Suffolk breeds established in North America to develop multi-breed synthetic populations. In 1972, the genetic base of the newly formed population was closed to any further introduction of new breeding material. 20 years of selection followed leading to the development of one meat-type terminal sire breed, the Canadian Arcott, and two fecund-type dam breeds, the Outaouais and Rideau Arcott. HISTORICAL DEVELOPMENT OF SHEEP GENE FLOW 209

However, introducing exotic breeds from temperate climates into the tropics and subtropics is usually unnecessary (Horst and Reh, 1999). Indigenous breeds have also been spread based on their improved traits. The Improved Awassi sheep from Israel is an outstanding example where through selection in purebreeding of the Awassi an improvement was achieved, and gene flow started as the breed gained importance in several countries of the Middle East (Iraq, Syria, Lebanon, Jordan, and Israel) (Horst and Reh, 1999). Details are given in the case study. An example for systematic crossbreeding in sheep, recorded in scientific literature, is the introduction of Blackhead Persian and Dorper rams into Maassai flocks in East Africa (King et al., 1984 according to Gatenby, 1986). The prolific Chios breed, found in Greece and West Turkey has been used in intensively managed milk production system for purebreeding and crossbreeding purposes in the Aegean region (Horst and Reh, 1999). The remarkable adaptability and reproductive performance of the Blackhead Persian in subtropical production environments has led to the spread of the breed to many African countries and to South America for purebreeding and crossbreeding purposes (Horst and Reh, 1999). Oya (1992) gives an example of a successful sheep improvement program for Djallonke sheep in Ivory Coast. In addition to the 10,000 ewes in the nucleus flock, the project monitors from 10,000-11,000 ewes in other flocks. On request, selected genetic materials in the form of rams, ewes, and fresh semen, were supplied to Togo. Hair sheep of West African and Central American origin have been used to improve West Malaysian and Sumatran wool breeds through crossbreeding, making use of the introduction of the hair coat and the reproductive performance of the hair sheep breeds (Gatenby et al., 1997a; 1997b; Gatenby et al., 1994; Gatenby, 1986; Schäfer, 1998). Horst and Reh (1999) stated that the Barbados Blackbelly breed, a hair sheep breed from Central America, has gained international interest due to their remarkable reproductive performance.

2.4 Influence of technology and global mobility on the dissemination of genetic material Artificial insemination is less successful in sheep than in cattle. It requires capital-intensive production systems and is therefore not widespread where sheep are maintained on extensive systems and receive little individual attention. In the former USSR and former GDR, where flocks were run on collective and state farms, AI was widely practised but it collapsed for cost reasons after the economic changes. In general, AI in sheep has had an important impact in schemes where the use of fresh semen is practical, such as today’s dairy sheep breeding programmes. According to the ICAR report of the working group on milk recording in sheep, in France 154,000 doses of fresh semen per year were used in 2001 alone in the breeding scheme of the Lacaune with 458 AI progeny tested rams per year. Italy and Spain also use AI with around 20,000 doses per year in their dairy sheep breeding schemes. (For further details see Annex 9.1 A 5.) Furthermore, the strategic use of laparoscopic intra-uterine AI of frozen semen, and in some cases cervical AI, to create genetic links across flocks, is central to the success of 210 CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW sheep sire referencing schemes. There is no doubt that new techniques, which produce high conception rates and use less invasive insemination methods than laparoscopy, could have a major impact on the dissemination of genetic improvement in the sheep industries of many countries (Simm, 1998).

3 CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW

To describe the current status of gene flow, the transfer routes of breeding materials needed to be identified and their impact assessed. The treatment envisaged was to analyse selected export and import data of breeding stock and semen, supplemented by the analysis of indicators of gene flow. Two indicators were investigated: the changes in national breed compositions and the influence of foreign genetic material in existing sheep breeds. During the process it became evident that this claim faced many limitations. Compiled export and import data on breeding material were only available for EU-15 countries in the form of the Eurostat statistic database. A distinction at breed level was not made which limited specific conclusions on gene flow significantly. To gain export and import data beyond this European database, national contacts were searched for. The FAO-based DAD-IS was used to identify national experts. More data was sourced from national breeding associations via their websites. Altogether, six associations and 21 experts from 12 countries were contacted. Only three useful contributions were returned, very little compared to the effort made. In many cases, no reply was received at all. In other cases, the information was not available or not freely made accessible. Against this background, extracting information from publications and in particular from the Country Reports gained importance. To obtain a regional balance, country reports from ten countries were used: Poland, Russia, Bulgaria and Romania for Eastern Europe, North America, Canada, Paraguay, and Uruguay for The Americas, Australia for Oceania and China for Asia. However, these sources should not be relied on totally as the procedure is very time consuming and the results remain fragmentary. Concrete figures on breeding material exports and imports were rarely found. However, the reports contained valuable gene flow information where foreign genetic material was introduced.

3.1 Gene flow indicated by selected export and import

3.1.1 EU-15 countries

In the EU-15 countries the sheep population accounted for 100 million head in 2003 according to FAO statistic database. The population trend has followed the profitability of sheep production during the last decades. It was considered stable until the end of the 1900s however it has decreased ever since due to the decline of profitability. The distribution of sheep within the EU-15 countries as depicted in Table 4 for the year 2003 show that the United Kingdom accounted for one third of the overall sheep population, followed by Spain, France, Greece, and Italy. CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW 211

Table 4: Distribution of sheep population in the EU-15 countries including export and import of breeding animals in 2003

EU-15 countries Sheep population Breeding sheep Breeding sheep (head) export (head) import (head) United Kingdom 35,700,000 7 24 Spain 23,800,000 57,491 3,154 France 9,200,000 1,592 - Greece 8,900,000 - 6,161 Italy 7,900,000 - 2,373 Portugal 5,500,000 2,498 - Ireland 4,800,000 - - Germany 2,700,000 415 119 Netherlands 1,200,000 6 - Sweden 448,000 - - Austria 304,000 40 37 Belgium 146,000 73 1,069 Denmark 143,000 - - Finland 67,400 - - Luxemburg 9,000 - -

Source: Eurostat (2004)

The Eurostat statistics available provide insights into the scope and development of European exchange of purebred breeding sheep (Figure 3). Exchange among EU-15 countries was much higher than exchange with other European countries, which ranged from 2.4% to 9.5% of total exports over the years. The main exporter was Spain, which has dominated exports since 1997. In 2003, Spain accounted for 92.5% of the exports of purebred sheep from EU-15 countries. Main destination countries were Portugal (68.8%), Italy (14.4%), Greece (12.6%), and France (4.2%). France and Germany traded more breeding stock with European countries outside the EU-15 territory. Exports to these countries often exceeded that among EU-15 countries but figures vary greatly between years. Export destinations during the last decade were mainly the Balkan region. France further exported sheep to Switzerland, Turkey, Morocco and Algeria, Venezuela, Brazil, China, and Japan. Germany exported to Switzerland, Brazil, China, and Indonesia. 212 CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW

Figure 3: Development of exports and imports of breeding sheep in EU-15 countries from 1992-2003

80

70

60

50

Exports 40 Imports

30 Live animals (thousands) 20

10

0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Imports into EU-15 countries were generally very small scale. If breeding stock was imported, imports came exclusively from European countries. Imports of breeding stock from countries outside the EU-15 territory during the last decade ranged from 0.03% to 3.8% of total imports. Suppliers were Switzerland, Poland, Chile, Argentina, and Uruguay. Differentiation of exchange of breeding stock at breed level was not possible. It may be concluded that EU-15 countries are net-exporters of purebred sheep with Spain playing a dominant role. Exchange takes place mainly among the EU-15 countries but Eastern Europe is important as an additional destination. Global exchanges also take place but are limited to individual contacts with single countries. The proportion of breeds of transferred purebred animals were not known and therefore no trends or reasons for transfers could be derived, nor conclusions drawn on their impact, for example on population composition.

3.1.2 Eastern Europe

No statistic database on exports or imports of breeding stock was available for Eastern European countries. Selected export and import figures were found only in the Russian country report. In this report, Russia states a gene influx for the sheep breeding sector and an intensive replacement of its genetic material. Export/import figures given in the report are summarised in Table 5 below. An orientation towards fine wool production is stated by crossbreeding local sheep with imported Australian, Mazevski and Siberian Merino. Crossbreeding with Rambouillet from the USA is also mentioned, but not supported by concrete figures. CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW 213

Table 5: Russian sheep export and import figures of sheep

Year Breed Country Number Imports 1990-1995 Australian from Australia, 226 males, Merino, Texel Finland 50 females 2001-2002 Mongolian from China 870 females Merino Exports 2001-2003 Romanovskaya, to Latvia, 193 females Tsigayskaya, Kyrgyzstan, Orenburgskaya Kazakhstan

Source: Country Report Russian Federation (2003)

3.1.3 America

This chapter concentrates on North and South America, neglecting Central American countries. Information was available through country reports and in the case of Argentina from personal communication (Mueller, 2004). For Canada, no information on import/ export of breeding animals could be extracted from its country report. In the USA, the sheep breeding sector experienced a major loss of profitability during the last half of the 20th century. Numbers of breeding ewes dropped about 40% during the last decade to 5 million head in the year 2000. Due to a growing interest in hair sheep breeds 2,165 Dorper sheep from South Africa were imported 1990 - 2000 (Annex 9.1 A 6). An import of 347 Texel sheep 1990 - 2000 could also be identified from the table. To develop niche markets, an increased interest in sheep milk production with imports of East Frisian Milk sheep and Lacaune as well as several imports from Canada and France are reported. However, numbers of imported animals were not given. In Uruguay, with the exception of the Merilin synthetic breed, all breeds utilised are of foreign origin and most are derived from the Merino breed. Current gene flow into the country consists of imports of breeding stock and semen with varying degrees of continuity. In the case of Merino and Corriedale the imports originate from Australia and New Zealand. The Ideal breed is imported from Australia solely and the Romney Marsh from New Zealand. However, concrete figures were not given. In Argentina, the current status of gene flow is an estimated 5,000 doses of frozen Merino semen and about 5 live Merino rams imported yearly, with about 95% from Australia and 5% from New Zealand. Exports of Merino breeding stock from Argentina are limited to neighbouring countries.

3.1.4 Oceania

Concrete figures on gene flow for the Oceania region were not found in the given sources. Although export/import statistics for Australia and New Zealand are kept by national breeding associations, they were not made freely accessible to the authors.

3.1.5 Asia

For Asia, China was selected to depict gene flow of a major Asian country. Information was extracted from its country report (Country Report of the Peoples Republic of China, 214 CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW

2003). No information more recent than 1989 was found. The reports states that before 1989 according to incomplete statistics cited in the country report, 25 sheep breeds were introduced mainly from Russia, France, the UK, Germany, the Netherlands, Denmark, the USA, Canada, Australia, and New Zealand.

3.1.6 Middle East

The case study on the Improved Awassi breed traced in detail the worldwide transfer of breeding material from the Improved Awassi and the Assaf breed. Information on the current gene flow from these breeds can also be obtained from the case study on the worldwide gene flow of the improved Awassi and Assaf breeds of sheep from Israel.

3.1.7 Africa

Obtaining information on gene flow from developing regions was expected to be extremely difficult. Export and import figures for developing countries were not available even for live animals through the FAO statistic database, and were certainly not available for breeding stock. South Africa was expected to provide a better information base compared to other African countries. However, a country report did not exist. The only information on current gene flow was contributed by the South African Dorper Society. With 6.6 million head in 2002, the synthetic Dorper as a single-purpose breed for mutton has become the second largest breed in South Africa, with an increasing population trend. According to information from the South African Dorper Society, the Dorper breed has been distributed to many other countries in the world, such as Australia, New Zealand, Canada, America, Mexico, Brazil, Switzerland, China, Botswana, Zimbabwe, Namibia. (Milne, 2004).

3.2 Gene flow indicated by the introduction of foreign genetic material Indicators that gene flow has taken place and the basis for impact analysis are information on changes in the national breed composition and on the introduction of foreign genetic material into existing breeds. Sheep production following World War II was characterised by economic pressure for producing cheep commodities to meet the growing demand of the increasing human population and improve productivity. In parallel, the wool market hit a crisis in 1968 and the wool price collapsed in 1989 - 1991. Due to the marginal role of sheep husbandry in national economy, sheep production suffered a dramatic decline of profitability. The last decade showed both a dramatic reduction of sheep population and the orientation from wool to meat production. This in general called for imports of exotic breeds with improved meat and prolificacy characteristics and an influx of foreign genetic material in many parts of the world. The orientation towards meat production increased interest in hair sheep breeds outside their traditional production environments. When shearing costs exceed the wool price, hair sheep breeds become more popular as they produce meat without needing shearing. Additionally, hair sheep as indigenous to the tropics are known for their non-seasonality, as well as high twinning rates, good mother abilities and resistance to particular diseases. High prolificacy has been reported particularly for the Barbados Blackbelly and the St. Croix hair breeds and for the Djallonke sheep. The Dorper breed is popular based on its CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW 215 growth performance. Therefore, as well as the sources mentioned in the introduction of the chapter, statements from publications on the particular movements of these hair sheep breeds have been used for this chapter.

3.2.1 Europe

Sheep production in Eastern and central Europe was additionally characterised by the introduction of the market economy in the early 1990s. Hungary provides a good example to prove the increasing importance of alternative breeds. In Hungary, virtually all sheep raised before 1968 were Merinos. The drastic fall in wool prices resulted in a switch to lamb fattening technologies at the beginning of the 1970s. Income proportion changed significantly from wool to meat. Additionally, changes in the political system, state subsidies, and the structure of ownership affected not only the development of the sheep population but also its composition. The current breed composition is shown in Table 6 below (Fésüs et al., 2002). The Merino breed had to give up its monopoly position. While today 86.9% belong to the Merino breed, the remaining stock is divided among small indigenous populations of animals specialised for meat and milking as well as prolific breeds (Fésüs et al., 2002). Before the 1960s, upgrading the Hungarian Merino was based on introducing Soviet Merino breeds including sheep from Caucasus, Stavropol, Groznyy, and most of all Askania. After the 1960s, when meat purpose breeding got under way, Merino Precoce from France and German Mutton Merino from East and West Germany were used for crossbreeding purpose. Merinos were crossed with Kent and Corriedale breeds. Crossing with Australian Merinos introduced the Booroola gene into Hungarian populations. For the dairy breed fraction it is known that Hungary has imported Awassi sheep from Israel.

Table 6: Breeds used in Merino breeding in the past and the recent Merino breeds in Hungary

Breeds used in the past to Recent Merino breeds in Distribution of breed groups improve the Hungarian Hungary in 2001 Merino Ascanian Merino Hungarian Merino Merinos 86.9% Groznyy Merino German Mutton Merino Mutton breeds 6.9% Stavropol Merino Merino Landschaf Dairy breeds 2.0% Caucasian Merino Booroola Merino Indigenous breeds 4.2% Merino Precoce Prolific Merino Mutton Merino of GDR Mutton Merino of FRG German Mutton Merino Australian Merino Booroola Merino

Source: Fésüs et al. (2002) 216 CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW

In Poland since the early 1980s the sheep population has grown steadily due to the increased export opportunities of prime lambs in Western Europe. This development demanded for a re-orientation from wool to meat production and resulted in importing many breeds, both meat and prolific ones for pure- and crossbreeding purposes. While in 1976 only 9 breeds had been stated, in 1986, when sheep numbers in Poland reached their peak, performance recording covered 26 different sheep populations including backcrosses. Until 1990, the population dropped slowly, then very quickly after the market economy was introduced and wool price collapsed. However, at the same time, the number of genotype groups increased, partly due to the new imports. Table 7 below gives details on 14,645 imported sheep of 12 identified breeds until 1990. At that time, these imported purebred sheep accounted for 4.1% of the sheep population. By 2000, these breed groups accounted for 15% of the total sheep population with 0.4% belonging to purebred Longwool breeds, 0.7% to prolific breeds and 13.9% to meat breeds, not including 2.3% backcrosses and 0.4% prolific crosses. Breed numbers were quoted with 33 breeds in 2000 (Polish Union of Sheep-Farmers, 1991). CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW 217

Table 7: Changes in breed structure in sheep population of Poland between 1976 and 2000

Breeds Number of ewes under recording scheme and their share in total active population 1976 1990 2000 Number % Number % Number % Polish Merino 158,255 43.5 171,241 47.8 31,988 29.8 German Merino 1,861 0.5 97 0.1 Polish Lowland In Wielkopolska type 52,184 14.3 22,982 6.4 4,388 4.1 In Lincoln type 5,550 1.5 - - - - In Corriedale type 5,986 1.7 277 0.2 Other Polish Lowland 57,348 15.7 87,933 24.5 26,035 24.3 Polish Longwool: In Pomorska type 39,949 11.0 15,969 4.5 5,760 5.4 In Suski type 16,594 4.6 - - - - In Kamieniecka type 12,922 3.6 1,625 1.5 In Pogorze type 4,874 1.4 1,755 1.6 Olkuska 50 0 -- Other Polish Longwool 17,104 4.7 9,520 2.6 779 0.7 Polish Mountain Sheep 2,086 2.6 5,152 1.4 11,306 10.5 Other purebred sheep 15,015 2.1 - - - - Polish Health Sheep 2,037 0.6 1,491 1.4 Swiniarka - 1630.2 Purebred imported sheep (Kent, Leine, East 14,645 4.1 - - Friesian, Finn, Lincoln, Booroola, Texel, Ile de France, Blackheaded Mutton Sheep, Suffolk, Berrichon du Cher, Dorset Purebred Longwool sheep (Kent, Leine) 431 0.4 Prolific sheep (East Friesian, Bergschaf, Finn, 758 0.7 Olkuska, Romanov) Meat sheep (Texel, Ile de France, Blackheaded 14,988 13.9 Mutton Sheep, Suffolk, Berrichon du Cher, Dorset Horn, Charolais, White Alpine) Synthetic line 874 0.2 2,627 2.4 Backcrosses 2,532 0.7 2,524 2.3 Prolific crosses 476 0.4 Total 364,085 100 358,578 100 107,468 100 Total population size (in thousands) 3,429.9 4,158.5 361.6 Including ewes 1,780.9 2,437.5 228.8 % of active population 20.4 14.7 48.2 Intensive replacement of the genetic material is also stated in the sheep breeding sector in Russia. Local sheep breeds are crossed with Merino from Australian and with Rambouillet from the USA. 218 CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW

For Europe, although not a tropical production environment, similar observations can be made. The introduction of Cameroon hair sheep into Germany had been reported by Gregg (1998). The introduction of the synthetic Dorper breed from South Africa in Germany in 1994 was described by Holtz (1998), when, scientifically accompanied and motivated, Dorper embryos and deep-frozen semen from some elite flocks in the north-eastern Cape Province and of the Oranje Freestate were introduced to Fichtenberg in Germany. Here, first purebred Dorper lambs were born in 1995. Some nucleus flocks were established in this area. Fahmy (1996) reported the transfer of the prolific Chios breed from Greece into many countries of the Near East

3.2.2 America

In the USA, the sheep breeding sector experienced a major loss of profitability during the last half of the 20th century. The number of breeding ewes dropped about 40% during the 1990s to 5 million heads in the year 2000. There are approximately 50 sheep breeds in the USA. Details are given in Annex 9.1 A 6. Of these Rambouillet, Suffolk, Hamshire, and Dorset are considered major breeds. The wool component has reduced over the last decade and about 80-95% of income was generated from lamb production in extensive management systems. Due to these market forces there is a growing interest in hair sheep breeds which is partly satisfied by importing the Dorper breed from South Africa. Numbers of imported animals extracted from the tables give 2,165 Dorper sheep from 1990 - 2000. In connection with the development of niche markets, an increased interest in sheep milk production is reported and imports of East Frisian Milk sheep and Lacaune as well as several imports from Canada and France are reported. However, numbers of imported animals were not given. For both the USA and Canada, Minhorst (1997) and Minhorst et al. (1999) reported the successful management of Barbados Blackbelly sheep and St. Croix flocks, both descended from Middle American breeds. St. Croix sheep were first introduced to the USA in the 1960s and were used to develop the Katahdin breed. Purebred flocks were kept in Utah, Florida, Mississippi, New York, Ohio, and California. In 1992, a small flock was exported to Canada. St. Croix sheep were also exported from Utah to Mexico and sold to producers for crossbreeding purposes with Pelibuey (Fahmy, 1996). However, the impact of these introductions is not known. In Uruguay, before the collapse of the wool price 1989-1991, crossbreeding the Criollo breed with Merinos from various sources has significantly improved wool quality and quantity. The genetic material was reported to come from Europe, USA, and Australia, and the wool characteristics changed from very thick, low quality to improved delicacy, colour, and smoothness. According to Uruguay’s country report crossbreeding with the Criollo led to almost total absorption. This period was followed by a predominance of crossings with Lincoln and Romney Marsh in order to meet the great demand for sheep meat in Europe. The Merilin breed was developed through crossing and selecting Merino (66% blood share) with Lincoln (33%). After 1930 the local genotypes were absorbed with synthetic breeds imported from Australia and New Zealand, mainly Corriedale and Ideal. From the 1990s, the relative importance of sheep meat increased and the interest for specialised breeds in meat production to be used in terminal cross systems arose. Imported breeds CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW 219 were Hamshire Down, Southdown, Suffolk, Ile de France, Texel and Poll Dorset of unquoted origin. Due to improved fine wool prices, the Dohne Merino breed from Australia has also been introduced to produce a double purpose meat and fine wool sheep. Therefore, with the exception of the Merilin synthetic breed in Uruguay, all breeds utilised are of foreign origin and those mostly utilised were derived from the Merino breed. According to Horst and Reh (1999) Corriedale crosses make a share of 50% of the population, 29% are crosses, and 10% are Merilin sheep. Current gene flow into the country is made up of imports of breeding stock and semen with varying degree of continuity. In the case of Merino and Corriedale the imports come from Australia and New Zealand. The Ideal breed is imported from Australia only and the Romney Marsh from New Zealand. In Paraguay sheep and goat stock numbers are increasing due to high prices of cattle meat. Sheep breeds are mostly Hamshire Down, Texel, Suffolk, Santa Ines, Criollo, Border Leicester, Corriedale, and Romney Marsh suggesting parallels with sheep population developments in Uruguay. However, information on the current status of gene flow was not found in the country report. In Argentina, total sheep population has declined dramatically from the early 1970s. Millions of sheep were replaced or killed, largely due to low wool prices and more profitable agriculture alternatives such as cropping and beef in favourable areas (pampas), but also due to natural disasters like volcano eruption (1992) and heavy snowfall in marginal areas such as Patagonia. The last decade has been more stable with some stock recovery due to improved fine wool prices. In the more favourable areas dual-purpose breeds are run, such as Corriedale originally from New Zealand, while the Merino is found in more marginal areas. Merinos remain in the southern cold Patagonia. The Argentine Merino population is about 6 million head, which corresponds to about 43% of the sheep population. The current status of gene flow is such that an estimated 5,000 doses of frozen semen and about 5 live rams are imported yearly to Argentina with about 95% from Australia and 5% from New Zealand. Merinos in Argentina are almost all descended from Australian Merinos, with over 50 years of continuous imports of stock from Australia. The general trend in breeding is towards finer wool and increased wool quality. Recently, superfine wool sheep from several studs were imported. There is some crossbreeding with meat-type breeds. Exports of Merino breeding stock from Argentina are limited to neighbouring countries (Mueller, 2004). Milne (2004) reported that the Dorper breed is popular in Mexico and Brazil. Fischer et al. (1999) report another example of hair sheep successfully introduced to the tropical region of South America. In 1990, 100 hair sheep of the breeds Barbados Blackbelly and Pelibuey-West African were imported to the Province Sucombios, Ecuador as part of the PROFOR project. In 1994, 170 animals followed. From 1994 onwards, the hair sheep production in this area could maintain itself without any additional imports and in 1998 had reached a population of about 1,000 head. The production performance of both breeds was reported to be comparable to that in other tropical and subtropical regions (Fischer et al., 1999). 220 CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW

3.2.3 Oceania

Australia has the second largest sheep population in the world, and is the largest live sheep exporter mainly for human consumption. Development of the Australian sheep sector has followed the changing profitability of wool production world wide: numbers in Australia reached a peak of 180 million in 1970. In general, numbers have fallen since then. Poor market prospects for wool after 1990 impacted the flock size and sheep numbers fell rapidly until 1995. The decline continued more slowly until 1999. As a result of the uncertainties in the wool market, many traditional Merino breeders have changed to alternative breeds which specialise in meat production (Padbury, 2002). In 2000, 95% of Australia’s sheep flock were purebred Merinos used for wool production (85.1%). 10.4% were first crosses used for prime lamb production using Border Leister sires (Simm, 1998). The remaining 4.5% represent more than 40 different breeds. Which breeds were used for the terminal crossing were not stated. Because of its strict quarantine system, it can be assumed that the breeds used were mainly not imported. However, Milne (2004) reports the transfer of Dorper breeding animals to Australia and New Zealand.

3.2.4 Asia

China, with 143 million sheep, keeps the world’s largest sheep population. Most important for sheep production systems are the pastoral zones of northwest China including Inner Mongolia, Gansu Xinjiang, Quinghai and where native breeds have adapted to local environments. Some exotic breeds were imported in order to improve production performance of these native breeds through crossbreeding. To improve fine-wool production a series of new breeds were developed including Chinese Merino, East-North Fine Wool and Aohan Fine wool. To improve the local sheep breeds′ meat production, meat purpose exotic sheep breeds such as Polled Dorset, Suffolk, Texel, and Charolais have been introduced to be used on small-tailed Han and Mongolian ewes. Altogether, before 1989 according to incomplete statistics in the country report, 25 sheep breeds were introduced mainly from Russia, France, the UK, Germany, the Netherlands, Denmark, the USA, Canada, Australia, and New Zealand. The introduction of Dorper breeding animals in China is confirmed by Milne (2004). Well documented is the introduction of hair sheep breeds into the tropical production environment of Southeast Asia in two scientific projects to develop a synthetic breed based on local breeds. For example, the Cameroon hair sheep from Germany was introduced into Malaysia for crossbreeding with local Longtail sheep (Schäfer, 1998). The DAD-IS indicates that hair sheep breeds (Barbados Blackbelly, Dorper, and Morada Nova) continue to be introduced into Malaysia for crossbreeding purposes. For another example, improved Barbados Blackbelly and St. Croix hair sheep from the USA were introduced into Indonesia for crossbreeding with Sumatran wool breeds (Gatenby et al., 1997a; 1997b; Gatenby et al., 1994; Gatenby, 1986). Information on the current impact of these breeds was not available.

3.2.5 Middle East

The case study on the worldwide gene flow of the improved Awassi and Assaf breeds of sheep from Israel depicts the flow of the Improved Awassi and the Assaf breed as prolific CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW 221 breeds from Israel into many countries worldwide. Detailed information on the introduction of these breeds in many countries can be obtained from there.

3.2.6 Africa

During the last decades, the South African sheep population has steadily decreased due to a national programme aimed at reducing . According to the FAO statistics, for the last decade there has been a fairly stable number of about 29 million sheep in South Africa, half of it the Merino breed. With 6.6 million head in 2002, the synthetic Dorper as a single-purpose breed for mutton has become the second largest breed in South Africa with an increasing population trend. According to the South African Dorper Society, the Dorper breed has been distributed to many other countries in the world, such as Australia, New Zealand, Canada, America, Mexico, Brazil, Switzerland, China, Botswana, Zimbabwe, Namibia and Swaziland. (Milne, 2004). According to Meyn (2005) Dorpers also came to Tanzania, Kenya and Zambia. A prolific breed, the Moroccan D’man, is exported in considerable numbers to Iraq (Fahmy, 1996).

3.2.7 The Booroola gene - An example for the introduction of foreign genetic material

The orientation towards meat production led to an explosion of interest in importing prolific genes in sheep breeds. Better availability and improved transport, as well as advances in AI and embryo transfer techniques, made it easy to move prolific breeds across the continents to develop new prolific breeds by upgrading local breeds. It has been realised that it is faster, easier and cheaper to improve sheep production by incorporating prolificacy from imported breeds using local base populations, than selecting these populations, although the importation of undesirable traits may also be encountered. Some countries have banned exports of prolific sheep and so these sheep are found only in their country of origin. However most countries have adopted an open policy to exporting their genetic resources, allowing prolific sheep to spread around the world. Where government or para-government agencies imported prolific sheep, these imports were documented. However, private companies also imported many animals, and most of these could not be accurately verified (Fahmy, 1996). The most widespread breeds are Finnsheep, Romanov, and Booroola Merino. The spread and use of the Booroola Merino was chosen to serve as an example based on the data compiled by Fahmy (1996). The globally important Merino breed was until then characterised by a low reproductive performance which limited the use of it as A-line in crossbreeding programmes for meat production (Horst und Reh, 1999). In the 1980s, the Booroola gene, a single gene with a dominant effect on litter size, was discovered in a Merino population in Australia. Wilson (1991) discovered that the high prolificacy in Booroola sheep is the result of a mutation in the BMPIB(ALK6) receptor. Turner (1982) suggested that the highly prolific Booroola Merinos can be traced back to an early Australian flock known to include prolific Bengal sheep, imported in 1792/93 from Calcutta. The description of the Garole breed given by Ghalsasi and Nimbkar (1993) seemed to fit into this picture. Evidence for this hypothesis was presented by Davis et al. (2002) who tested prolific sheep from eight countries for the presence of the BMPIB 222 CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW receptor mutation related to the Booroola gene and confirmed its presence in Garole and Javanese Thintail sheep. Europe imported Booroola sheep initially from Australia. However, as a result of a moratorium on exporting live Booroola sheep from Australia, Booroola sheep were also sourced from other countries. New Zealand gained importance as 3 Booroola Merino rams had been donated by the Australian CSIRO before the discovery that high prolificacy in Booroola Merinos is controlled by a single gene. 74 rams and 34 ewes were imported by private owners, making New Zealand the major country of origin for purebred Booroola Merino breeding stock apart from Australia. Table 8 below summarises the spread of the Booroola Merino in Europe. Recently, Spain imported Booroola Merino sheep and crossed them with Sagurina and Spanish Merino. However, information on the country of origin was not found.

Table 8: Source and year of importation of Booroola gene into European countries

Country Source Years of importation Hungary New Zealand, Australia 1980, 1982, 1984, 1986, 1989 France Australia, New Zealand, UK 1981, 1982, 1984 UK Australia, New Zealand 1981, 1982, 1984 Czechoslovakia New Zealand 1985 The Netherlands UK 1986 Germany New Zealand, Hungary 1988, 1989, 1992, 1993 Poland New Zealand 1988

Source: Fahmy (1996)

Where purebred Booroola Merinos were available, the main intention was to establish a nucleus purebred flock and introduce the Booroola gene from this flock into various local Merino breeds until the Booroola gene was homozygous throughout the local Merino population. However, after the Australian moratorium, Booroola gene imports were increasingly based on Booroola Merinos crossbred with various breeds. 22 Booroola Merino rams were imported from Australia and New Zealand in 1980 with Hungarian Merino sheep. The crossbreds became important to introduce the gene into German Mutton Merino. In America, research stations in the USA and Canada imported purebred homozygous Booroola Merino rams and embryos to establish purebred flocks. As well as a private USA company, which imported purebred homozygous Booroola Merinos from New Zealand, crossbred homozygous animals were imported by private sheep breeders in the USA. In Canada, many farms bought rams for slaughter, which were heterozygous for the Booroola gene and used them for breeding. Uruguay, Brazil and Chile were the only countries in South America to import Booroola Merino genes, exceptionally in the form of rams. Except for Uruguay, where 6 farms housing 6,000 sheep are using carrier animals, the Booroola gene’s influence on South American sheep populations remains limited. The only two imports of Booroola to tropical and subtropical countries in Africa and Asia were to Israel from New Zealand for crossbreeding purposes and to South Africa from CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW 223

Australia. In Israel the imports had a very large influence on breeding and led to the development of the Afec Awassi and Affec Assaf lines. Figure 4 illustrates the flow of the Booroola gene based on the information compiled by Fahmy (1996).

Figure 4: Geographic illustration of the flow of the Booroola gene

High volume Europe Low volume United Kingdom Suggested

Netherlands Poland North America France Czech Rep. Canada Germany Hungary USA Spain (Garole Asia breed) Middle East India Brazil Israel 18th (Afec Awassi century Afec Assaf) Chile Indonesia (Javanese Thintail) Uruguay South America

Africa Australia South Africa Oceania (Booroola Merino) New Zealand

Source: own elaboration, data from Fahmy (1996)

3.3 Gene flow indicated by Country Report excerpts The analysed excerpts from the Country Reports revealed little impact of sheep gene flow in most countries. Although sheep imports were reported by many countries, only North America and Western and Northern Europe also reported exports from their countries. However, conclusions from the summaries given in Table 9 cannot be drawn, because major players in the sheep industry are not included, like Australian, New Zealand or France. Narrative extracts from single country report excerpts are summarised in Annex 9.1 A 7. They also reflect a very heterogeneous state of information, serve illustrative purposes and should not be utilised to draw conclusions.

224 CURRENT STATUS AND ACTUAL TRENDS OF SHEEP GENE FLOW

Table 9: Country Report excerpts on gene flow of sheep

Breed Extent of gene flow Main breeds Breed Disadvantages Country report excerpts in and out of the imported replace- of exotic breeds containing information region ment In Out Africa low - 0 Dorper, Merino Low Not named Botswana, Ethiopia, medium Uganda, Zambia Eastern and low - 0 Merino low Poorly adapted Albania Bosnia- Southern medium to local breeding Herzegovina, Hungary, Europe systems, pests Romania, Russian and diseases Federation, Serbia and Montenegro, Slovenia Latin low 0 Criollo medium Not named El Salvador America North medium medium Suffolk low Not named Canada, United States of America America Oceania none Southern medium 0 Merino medium Replacement of Bhutan, China, Indonesia, and Eastern - high local breeds Japan, Kyrgyzstan, Nepal, Asia Pakistan, Philippines, Tajikistan, Vietnam Western and low low Charolais, Texel low Not named Ireland, Latvia, Sweden, Northern United Kingdom Europe Western low 0 American low Not named Armenia Asia Rambouillet, English Lincoln REFERENCES GLOBAL GENE FLOW OF SHEEP 225

4 REFERENCES GLOBAL GENE FLOW OF SHEEP

Astruc, J.M., F. Barrillet, M. Fioretti, M. Gabina, E. Gootwine, A.P. Mavrogenis, F.J. Romberg, S.R. Sanna, and E. Stefanake. 2004. Report of the working group on milk recording of sheep. ICAR Report. Results available on the ICAR website www.icar.org (01.06.2005). Baraja, F. 2002. Merino breed in Spain and its production. 6th Merino World Conference, Budapest, Hungary. Country Report of the Russian Federation for FAO’s State of the World’s Animal Genetic Resources. 2003. Ministry of Agriculture of the Russian Federation. DAD-IS. 2004. FAO information system for the Global Strategy for the Management of Farm Animal Genetic Resources. http://www.fao.org/dad-is. (01.12.2004) Davis, H.D., S.M. Galloway, I.K. Ross, S.M. Gregan, J. Ward, B.V. Nimbkar, P.M. Ghalsasi, C. Nimbkar, G.D. Gray, Subandriyo, I. Inounu, B. Tiesnamurti, E. Martyniuk, E. Eythorsdottir, P. Mulsant, F. Lecerf, J.P. Hanrahan, G.E. Bradford, and T. Wilson. 2002. DNA tests in prolific sheep from eight countries provide new evidence on origin of the Booroola (FecB) mutation. Biology of Reproduction 66:1869-1874. De Waal, H.O. and W.J. Combrink. 2000. The development of the Dorper, its nutrition and a perspective of the grazing ruminant on velt. Small Ruminant Research 36:103-117. Dmitriev, N.G. and L.K. Ernst (eds.). 1989. Animal Genetic Resources of the USSR. FAO Animal Production and Health Paper 65. FAO, Rome, Italy. Epstein, H. and I.L. Mason. 1971. The origin of the domestic animals of Africa, Volume II. Africana Publishing Corporation, New York, USA. Eurostat. 2004. EUROSTAT-COMEX Databank. European Union, Luxemburg. Fahmy, M.H. 1996. Prolific sheep. CAB International, Wallingford, UK. FAO. 2004. Global Livestock Production and Health Atlas (GLiPHA). 1[1]. 2004. FAO, Rome, Italy. Fésüs, L., L. Sáfár, P. Hajduk, and p. Székely. 2002. The determining role of Merino in Hungarian sheep breeding. 6th Merino World Conference, Budapest, Hungary. Fischer, J., C. Claus, A. Herrera, and G. Rahmann. 1999. Bedeutung der Haarschafhaltung für eine nachhaltige Nutzung der Regenwaldrandgebiete Südamerikas. GTZ, Tropenökologisches Begleitprogramm (TÖB), Seriennummer: TÖB F-V/9. Eschborn, Germany. Fitzhugh, H.A. and G.E. Bradford. 1983 Hair sheep of Western Africa and the Americas - A genetic resource for the tropics, Westview Press, Boulder, Colorado, USA. Gatenby, R.M. 1986. Sheep production in the tropics and subtropics. Longman, London and New York. 226 REFERENCES GLOBAL GENE FLOW OF SHEEP

Gatenby, R.M., M. Doloksaribu, and E. Romjali. 1994. The potential of hair sheep for the humid tropics of Southeast Asia. 7th Asian-Australian Association for Animal Production Societies (AAAP) Congress, 11-16 July 1994, Bali, Indonesia. 25-38. Gatenby, R.M., G.E. Bradford, M. Doloksaribur, E. Romjali, A.D. Pitono, and H. Sakuhl. 1997a. Comparison of Sumatra sheep and three hair sheep crossbreds. I. Growth, mortality and wool cover of F1-lamb. Small Ruminant Research 25:1-7. Gatenby, R.M., M. Doloksaribu, G.E. Bradford, E. Romjali, A. Baubara, and I. Mirza. 1997b. Comparison of Sumatra sheep and three hair sheep corssbreds. II: Reproductive performance of F1-ewes. Small Ruminant Research 25:161-167. Ghalalsasi, P.M.G.A. and B.V. Nimbkar 1993. The Garole - microsheep of Bengal, India. Animal Genetic Resources Information 12:73-79; cited by: Fahmy M.H. 1996. Handbook of Australian Livestock, Australian Meat and Livestock Corporation, 1989. 3rd edition. Haring, F. 1984. Schafzucht. Ulmer Verlag, Stuttgart, Germany. Henseler, H. 1944. Abstammungs- und Entwicklungsgeschichte des Schafes. In: Doehner, H. Handbuch der Schafzucht. Bd. I, Parey, Berlin. Henson E.L. 1992. In situ conservation of livestock and poultry. FAO Animal Production and Health Paper 99. FAO, Rome, Italy. Holtz, W. 1998. Keine Schur? Das Dorper-Schaf hält Einzug in Deutschland. Deutsche Schafzucht 7:154-156 Horst, P. and I. Reh. 1999. Tierzucht in den Tropen und Subtropen. Handbuch der Landwirtschaft und Ernährung in den Entwicklungsländern, Ulmer Verlag, Stuttgart, Germany. King, J.M., A.R. Sayers, C.P. Peacock, and E. Kontrohr. 1984. Maasai herd and flock structures in relation to livestock wealth, climate and development. Agricultural Systems 13: 21-56; cited by Gatenby, 1986. Lauvergne J.J. 1990. Endangered breeds of sheep. In: Animal genetic resources. Wiener, G. (ed.) FAO Animal Production and Health Paper 80. FAO, Rome, Italy. Ludwig, A. 1997. Abstammung der europäischen Hausschafe und Phylogenie der eurasischen Wildschafe. Dissertation, Humboldt-Universität, Berlin, Germany. Meyn, K. 2005. Personal communication. Milne, C. 2000. The history of the Dorper sheep. Small Ruminant Research 36:99-102. Milne, C. 2004. Personal communication. Minhorst, R. 1997. Nolana - für Fleischproduktion und Landschaftspflege. Deutsche Schafzucht 23: 536-538. Minhorst R., H. Lang and P. Stockdale. 1999. Haarschafe, Wald und Wild als integriertes System. Deutsche Schafzucht 22: 540-543. Mueller, J. 2004. Personal communication. Muigai, A.W.T., J. Hirbo, S. Sharkey, J.E.O. Rege, H. Blackburn and O. Honotte. 2002. Genetic diversity and relationships of hair sheep breeds of the Americas: First REFERENCES GLOBAL GENE FLOW OF SHEEP 227

Results. 7th World Congress on Genetic Applied to Livestock Production, 19- 23August 2002, Montpellier, France. New Zealand Merino Breeders Association. 2005. http://www.nzsheep.co.nz (18.12.2004). Oya, A.B. 1992. The national sheep improvement programme in Cote d’Ivoire. In: Rege, J.E.O. and M.E. Lipner (eds.) African animal genetic resources: Their characterisation, conservation, and utilisation. Proceedings of the Research Planning Workshop held at ILCA, Addis Ababa, Ethiopia, 19-21 February. Padbury, T.A.D. 2002. The Australian Merino industry (1788-2002). 6th Merino World Conference, Budapest, Hungary. Payne, W.J.A. and J. Hodges. 1997. Tropical cattle: origins, breeds, and breeding policies. 1st Edition. Blackwell Science, Oxford. Payne, W.J.A. and R.T. Wilson. 1999. An introduction to animal husbandry in the tropics. 5th Edition. Blackwell Science, Oxford. Polish Union of Sheep-Farmers. 1991. Sheep and goat breeding in Poland in 1990. Warsaw, 1991. Ponzoni, R.W. 1992. A global review of the genetic resources of sheep and goat breeds. In: Hodges, J. (ed.) The management of global animal genetic resources. FAO Animal Production and Health Paper 104. FAO, Rome, Italy. Ryder, M.L. 1984. Sheep. In: Carlson, I.L. (ed.) Evolution of domesticated animals. Longman, London and New York. Schäfer, C. 1998. The influence of crossbreeding Cameroon hair sheep with West Malaysian wool sheep on coat characteristics. Dissertation, Humboldt-Universität, Berlin, Germany. Scherf, B. 2000. World Watch List for Domestic Animal Diversity. 3rd Edition. FAO publication. http://dad.fao.org/en/refer/library/wwl/wwl3.pdf (01.06.2005). Schmidt, H. 1995. Zuchttierexport: Chinesen setzen auf Merinofleisch. Deutsche Schafzucht 6:135. Shelton, M. and E.A.P. Figueiredo. 1990. Hair sheep production in tropical and subtropical regions. Small Ruminant Collaborative Research Support Program, Davis, USA. Shrestha, J.N.B. 2005. Conserving domestic animal diversity among composite populations. Small Ruminant Research 56: 3-20. Simm, G. 1998. Genetic Improvement of Cattle and Sheep. Farming Press, Ipswich, UK. Strittmatter, K. 2003. Schafzucht. Ulmer Verlag, Stuttgart, Germany. Turner, H.N. 1982. Origins of the CSIRO Booroola. In: Piper L.R., B.M. Bindon, and R.D. Nethery (eds.) The Booroola Merino. CSIRO, Australia; cited by Fahmy, M.H. 1996. Wilson, T.R. 1991. Small ruminant production and the small ruminant genetic resource in tropical Africa. FAO Animal Production and Health Paper 88, FAO, Rome, Italy. Zeuner, F.E. 1963. A history of domesticated animals. Hutchinson, London.

228 REFERENCES GLOBAL GENE FLOW OF SHEEP

ANNEX 229

ANNEX

9.3 Global gene flow of goats

E. Alandia Robles, C. Gall, A. Valle Zárate

230 TABLE OF CONTENTS

TABLE OF CONTENTS

List of tables 230

List of figures 230

Abbreviations 230

1 Introduction 231

2 Historical development of goat gene flow 231 2.1 Origin and early distribution of the goat species 231 2.2 Influence of breeding methods, technology and global mobility on the spread of goats 233 2.3 Development and distribution of Angora goats 234 2.4 Development and distribution of Boer goats 236

3 Current status of goat gene flow 237 3.1 Goat gene flow indicated by Country Report excerpts 237

4 References global gene flow of goats 239

LIST OF TABLES

Table 1: Country Reports on gene flow of goats 238

LIST OF FIGURES

Figure 1: Probable spread of domestic goats from initial domestication (closed circle) to East and South Asia 231 Figure 2: Distribution of goat types 232

ABBREVIATIONS

AI Artificial insemination e.g. for example FAO Food and Agriculture Organisation of the United Nations USA United States of America USSR former Union of Socialist Soviet Republics ZASV Zoological Acclimatisation Society of Victory

INTRODUCTION 231

1 INTRODUCTION

Global gene flow of goats refers to the historic development of gene flow and the current status of gene flow for this species. The historic development concentrates on the influence of domestication, breed formation, human migration, and breeding methods on the diversification of goats. The evaluation of the current status of gene flow was exclusively based on Country Reports due to the lack of other reliable information sources.

2 HISTORICAL DEVELOPMENT OF GOAT GENE FLOW

2.1 Origin and early distribution of the goat species Goats belong to the genus Capra, with several species: C. ibex, C. pyrenaica (the Spanish wild goat), C. caucasia (the tur), C. aegagrus (Bezoar goat), C. falconeri (Markhor). The domestic goat (C. hircus) is believed to have a monophyletic root in C. aegagrus (Horst and Husain, 1991; Mason, 1981), but the Markhor and Ibex may also have contributed. (Nozawa, 1983). Figure 1 shows the probable spread of domestic goats through the Asian continent.

Figure 1: Probable spread of domestic goats from initial domestication (closed circle) to East and South Asia

Source: Devendra and Nozawa (1976) 232 HISTORICAL DEVELOPMENT OF GOAT GENE FLOW

Gradually, different goat types evolved in different areas, formed as much by the demand of man as by the environment. Goats with short ears are predominant in Europe, in North and coastal West Africa, in part of East Africa and in Southeast Asia. In Central Asia the main types are coarse-haired or cashmere producing, with short ears and twisted horns; they include the Angora that produces mohair fibre. Lop-eared goats are dominant in the Near East, parts of West and East Africa, southern Africa and parts of the Indian subcontinent. For the global distribution of goat types see Figure 2.

Figure 2: Distribution of goat types

(1) milk goats, (2) meat goats, (3) Angora goats, (4) Kashmir goats

Source: adapted from Künzi and Stranzinger (1993)

The number of goat breeds is estimated at approximately 300 by Horst and Husain (1991) and Pashaa and Saithanoob (2000), whereas Scherf (2000) reported 594 goat breeds currently recorded worldwide. The majority of goat breeds are found in the tropics and subtropics. However, breeds are difficult to identify in many parts of the world. While many populations are called certain breeds by their owners their genetic and phenotypic characteristics are not documented. For most of these populations phenotypical characterisation based on field records and herdbooks are lacking. In addition, effective reproductive isolation is not always given. In fact it is difficult to ascertain whether they are distinct breeds. As from the 19th century onwards, exotic goats were introduced in several European countries to create new breeds. Breed societies and herdbooks were founded in the 19th and early 20th century in England, and purebred goats were exported to many countries all over the world as the founder stock for establishing new breeds and for crossbreeding to improve the performance of indigenous goats. The prick-eared native goats of Britain were crossed with milk goats introduced from Switzerland and Nubian-type goats from Africa and India. The newly established Anglo Nubian with morphological characteristics of the Nubian-type goat was later distributed HISTORICAL DEVELOPMENT OF GOAT GENE FLOW 233 worldwide as a well-performing dairy breed, to tropical and subtropical countries in particular (Nozawa, 1983). The development of this breed and its spread is discussed in the case study "History and worldwide development of Anglo Nubian goats and their impacts in smallholder farms in Bolivia". Goats were not present in America and Oceania when European settlers arrived. West European goats were introduced from colonisation onwards: to North America from France and England, to Latin America from Spain and Portugal and to Australia from England (Porter, 1996). The Mexican Criollo goat is a mixture of the original Spanish and Portuguese breeds. It is predominantly Murciano-Granadina in the central region of Mexico, Alpine and Toggenburg in the Northern region and some Celtiberic White in Puebla (Glimp, 1995; Montaldo and Meza, 2000). When slaves were brought from West Africa to West Indian Islands they brought along African dwarf goats, and the West African hair goat. West African goats are considered to have contributed to the gene pool of American common goats (Nozawa, 1983). Some fine, generally large-framed dairy breeds were developed in India. The Damascus goat is a dairy breed of Western Asia. It was recently improved by the Agricultural Research Institute, Ministry of Agriculture and Natural Resource in Cyprus. It has gained international recognition as outstanding dairy breed for tropical and subtropical countries. While the herd is quite small much of this genetic resource has spread around the Mediterranean basin. In Africa south of the Zambezi River, goats were introduced late, shortly before and after the arrival of European settlers. The goat population is, therefore, derived from various breeds brought by Bantu tribes from the north (Gall, 1996). Feral goats are an important gene reserve. They originate from goats brought by the early explorers to ocean islands and left there for future meat supply. In Australia Feral goats stem from goats used during the 19th century for producing fibre (Mohair and Cashmere). With the arrival of fine wool sheep, goats were abandoned and some turned feral. At the end of the 20th century their number was estimated at 4 million.

2.2 Influence of breeding methods, technology and global mobility on the spread of goats In many countries, crossbreeding of local goat populations with bucks of specialised breeds was the main tool for genetic improvement. Purebred animals or semen were imported and used directly, or after crossbreeding and multiplying in local herds. Crossbreeding strategies favoured creating new breeds. The outstanding example is the development of dairy breeds of Central Europe. Most of these highly productive breeds were derived from local goats by upgrading with dairy breeds of Swiss origin. Even in those breeds that were developed mainly by within-breed selection, occasional introductions maintained a steady gene flow. Some new breeds were even formed mainly by importing pure stock or absorptive crossing (e.g. British Saanen or the Japanese Saanen). In Mexico, the widespread use for many years of sires from several specialised breeds imported from the USA, such as Anglo Nubian, Alpine, Saanen and Toggenburg, makes the current Criollo a multi-breed population composed mainly of Nubian crosses in herds 234 HISTORICAL DEVELOPMENT OF GOAT GENE FLOW used for meat production, and crosses of Granadina and breeds of alpine origin in milk producing herds (Montaldo and Meza, 2000). Similarly, albeit on a smaller scale, Criollos were upgraded in semiarid northern Venezuela and other Latin American countries. In South Africa local Boer ewes were crossbred since 1857 with Angora rams in order to upgrade for fibre production as quickly as possible (Porter, 1996). Between 1838 and 1896 about 3,000 Angoras were imported from Turkey. In 1906 the first Angora herdbook was established and in 1921 the South African Angora Breeding Association was founded (Gall, 2001; Nozawa, 1983; Porter, 1996). Further examples of breed formation through crossbreeding are the Indian Mohair in Southern India in 1973, created from a cross of 7/8 Angora and Sangamneri (Porter, 1996); and the Pafuri goat, created in Southwest Mozambique after the introduction in 1982 of South African Boer male goats and their crossing with Landim females (Gall, 1996). Crossbreeding was also used when introduced exotic breeds failed to adapt well to local conditions and/or requirements. In 1937 700 Angora goats were imported from the USA to the former USSR. Since the animals had difficulties in acclimatising, absorptive crossbreeding with local hair goats was started on state farms in Kazakhstan, Uzbekistan, Turkmenistan and Kirgiziya. Progeny from the second backcross to Angora has been bred inter-se since 1947. Selection was for fleece characteristics, body weight and carcass quality and the result was the development of the Soviet Mohair breed (Gall, 1996; Porter, 1996). The development of biotechnology (artificial insemination and deep-frozen semen conservation, embryo transfer) enhanced global mobility in the 20th century. It facilitated dissemination of breeds around the world and the establishment of new breeds. Breed improvement very often makes use of genetic resources obtained from outside the breeding area by crossbreeding thus inciting gene flow. This has happened in most improved breeds, fibre and dairy breeds in particular.

2.3 Development and distribution of Angora goats Angora goats have been developed since ancient times; their history is said to be older than the written records of man (Tuncel, 1987). The ancestor of the Angora goat is considered to be the Wild Goat of Persia, C. aegagrus but archaeological investigations suggest that the Angora goat may have descended from the wild goat C. prisca (Tuncel, 1987). Goats have been bred in Turkey for more than 4,000 years. They probably came originally from the east, possibly central China, Tibet or Turkistan. Goats, along with their herders, were driven to Turkey by Genghis Khan (Porter, 1996). Others claim that Central Anatolia, in Turkey, are their area of origin (Gall, 1996). The formation of industrially usable fibre (Mohair and Cashmere) is caused by specific genes. The interest in production of these fibres has prompted the flow of these genes from their Asian countries of origin to all other continents. During the 19th century in Australia suitable fibre was produced by goats. The development of the Australian Angora industry is a good example of the importance of commercial interest for the promotion of gene flow and the development of new breeds. The first Australian imports of Angora goats was from France in the 1830s. In 1856 a Melbourne merchant imported seven mohair goats acquired by the Zoological HISTORICAL DEVELOPMENT OF GOAT GENE FLOW 235

Acclimatisation Society of Victory and these were kept in Melbourne’s Royal Park. In the following years, France’s Imperial Acclimatisation Society presented ZASV with twelve top-class pure Angora from Asia Minor. More Asia Minor imports followed in 1869 and from 1871 to 1873, and by 1895 there were two thousand in the herd. Mohair productions became quite popular but with the advent of merino sheep breeding in the early twentieth century goats were abandoned. The Angora became almost extinct in Australia and by 1947 there was only one flock left (Gall, 1996). Many goats (Mohair- and Cashmere- bearing) went feral. The Angora bounced back during the 1970s, when wool prices were low and mohair fibre was in short supply internationally. Crossings of feral goats which had the genetic potential of producing fibre with Angora goats imported from South Africa and Texas, led within five generations to the development of a white Angora goat with good mohair quality. The Angora breeding started in 1970 with about 1,000 animals. Two breeding organisations were created and in 1988 there were 320,000 Angoras (Porter, 1996). In the 1970s China (previously a major source of cashmere) adopted a policy of processing its own fibre, forcing several processors in Britain, USA and Italy to look to Australia and New Zealand for alternative suppliers. In 1980 the processors Dawson International, particularly interested in white cashmere, moved to Australia to establish a goat farming industry. In New Zealand goats (some Angoras and a few cashmere) were introduced from 1842 onwards by acclimatisation societies on 22 offshore islands. The early Angora imports failed because they were susceptible to footrot and other diseases. Traces of Angora ancestry remained in the herds of Hawkes Bay and North Auckland (Mason, 1981). Many of the abandoned animals went feral. During 70-80 years in the bush they have adapted to New Zealand conditions, but fleece weight has declined. With the revival of the mohair industry in the 1970s, feral animals were used to upgrade Angoras for mohair production (Mason, 1981). In 1979 about 27,000 feral does were transported to South Island to be mated with Angora bucks. New Zealand rapidly expanded its fibre industry based on feral and Angoras. In 1984 there were about 800 Angora herds in the country; two years later there were some 2,000. In 1985 New Zealand produced 6 million kilograms of mohair (Porter, 1996). In order to build up a mohair industry in the USA, Angora goats (seven does and two bucks) were imported to Texas in 1849. Despite Turkey’s ban of Angora exports in 1881 another 300 goats went to the USA. In 1898, the Angora population was about 250,000 and a breeder’s society was established in this country in 1900 (Gall, 2001). South Africa started Angoras with very few animal brought from Turkey in 1838 and several major imports followed in the second half of the 19th century. By crossing with white Boer goats the Angora population was built up to about 3 million around 1900. By 1911 South Africa produced around 60% of the world Mohair (Gall, 2001) with 4.5 million Angora goats. A small Angora goat population exists in Lesotho, which is mostly the result of crossbreeding with South African stock. Some other countries like Pakistan and Madagascar have tried to establish Angoras (Horst and Husain, 1991) with variable success. A major herd exists in Latin America (Argentina). 236 HISTORICAL DEVELOPMENT OF GOAT GENE FLOW

The Soviet Mohair goat, officially recognised as a distinct breed in 1962, was developed from Asian fibre goats crossed to Angora goats from Turkey and in the 1930s from the USA (Gall, 1996). Angora goats spread all over the world from Turkey, and after Turkey’s export ban in 1881 were redistributed from other countries. Although detailed statistical records are not available it can be concluded that small numbers of animals were enough to establish a broad mohair industry in some countries. The existence of Angora goats in Western Europe was reported in the 16th century (1554). Angora goats were sent to France and in 1598 they were reported in Cyprus. In 1725, Dutch people tried, without success, to establish longhair Persian goats in South Africa. During the 18th century, Angora goats were introduced into Holland, England, Italy and France in small numbers but due to adaptation problems they perished (Arbiza, 1983; Horst and Husain, 1991).

2.4 Development and distribution of Boer goats In the last decades, goat meat demand has increased remarkably due to the economic development with improvement of living standards or, in countries like USA and Canada, due to the growing ethnic groups who prefer goat meat in their diet (Hispanic, African, Asian, Middle-Eastern and Caribbean people) (Luginbuhl, 2003). There is a rewarding market for goat meat in West-Asian Arab countries. As a consequence demand increased for animals with meat conformation. The history of the Improved Boer goat is another example of the development of breeds as a result of human demand. The Boer is a meat goat, developed in South Africa (East Cape) from the 18th century onward by selecting various indigenous goats. These are the spotted Hottentot goat, which possibly had its origins in the Nubian types that were said to have accompanied tribes during their southward migration down the east coast of Africa, and the Bantu goat. They were crossed with Indian goats and with high-class European dairy breeds (Gall, 1996; Malan, 2000; Porter, 1996). Polledness indicates the possible influence of European dairy goats (Casey and van Niekerk, 1988; Porter, 1996). Since the 1920s the Boer goat has been selectively bred for meat conformation. It is considered to be one of the most suitable goat breeds for meat production due to its body conformation, growth rate and carcass quality, adapts well to tropical conditions and has been introduced and crossbred with low-yielding goats in several parts of the world (Erasmus, 2000; Lu, 2002; Yonghong, 2001). Boer goats were introduced into Germany in 1980 in order to establish a meat goat by crossing with German dairy breeds. A herdbook for the Boer goat was formed in 1980 and, since 1985, repeated imports of frozen semen as well as frozen embryos in 1994, led to the establishment of a population of about 1,500 registered goats, distributed all over Germany (Gall, 1996; Porter, 1996; Sohnrey, 2000). In 1982, the German Boer type was introduced into Sri Lanka and crossed with local goats to improve meat production (Porter, 1996) and in 1988 the German Boer goat was first exported to Britain, where a Boer breed society was formed. Nowadays, breeding stock and frozen embryos are sold to breeders in several European countries and abroad (like Brazil, China and Israel) (Sohnrey, 2000). The Boer goat has also been widely used to increase meat production in Australia in the early 1990s. This was mainly a reaction on the market opportunities in the Arabian countries of Western Asia and the slump of wool prices prompting growers to change from wool to meat production. In 1989 the first of CURRENT STATUS OF GOAT GENE FLOW 237 hundreds of Boer goats were imported into Australia as live animals or as embryos. By the end of 1998 there were just under 10,000 registered pure Boer goats but hundreds of thousands of Boer crosses, mostly the result of mating between Boer bucks and feral does. From the mid 1990s onwards goat meat producing enterprises and Boer breeding developed, as an increasing number of traditional sheep and wool producers changed their livestock from sheep to goats (Murray et al., 2000). In the USA the demand for meat goats was related to the decline of the Mohair industry and a growing market for goat meat. Embryo transfer technology and AI has allowed the numbers of purebred Boer goats to increase from a few dozen to thousands in a couple of years (Cutrer, 1995). The first South African frozen Boer embryos were introduced in 1993. A Meat Goat Association was created in 1993 (Luginbuhl, 2003; Machen, 1997) and since 1995, the quality of Boer goats in the USA has improved remarkably due to sustained direct imports of live goats, semen or embryos. Up until 1998, approximately 4,000 Boer goats from South Africa were either imported directly from South Africa or via Australia or New Zealand. At present there are approximately 8,000 Boer goats in the USA and the meat goat industry is the fastest growing animal industry in the country (Malan, 2000; Sahlu, 2000). There are approximately 800,000 meat goats in the USA of which about 700,000 are in Texas. With the increased interest in meat production, several Asian countries (Indonesia, Malaysia, Singapore) have imported Boer goats (Malan, 2000). In China the Boer breed was introduced by artificial insemination to improve local goats for higher meat production. In 1995, for the first time, China imported 25 breeding Boer goats (8 males and 17 females) from Germany, and pelleted frozen semen was prepared in the fall of the same year. In 1995-97 more than 6,000 local does were inseminated with frozen and fresh semen. Following these first imports, about 3,000 have been introduced successively from Germany, South Africa, New Zealand and Australia. By 2000 the population of purebred Boer goats had reached 6,000 and was spread all over China (Gangyi et al., 2000; Yonghong, 2001). The number of Boer crossbreds totalled 400,000 in 1999, and increased to about 600,000 in 2001 (Yonghong, 2001). Today, China is exporting Boer goats.

3 CURRENT STATUS OF GOAT GENE FLOW

Information on current gene flow in goats is very scarce and has been sourced mainly from Country Reports. The valuable contribution from experts were mainly of analytic character and are comprised in the respective subchapters of the main findings.

3.1 Goat gene flow indicated by Country Report excerpts Information in Country Report excerpts regarding goat transfers are less extensive compared to other species. Many excerpts did not contain any mention of goats at all, so that a global conclusion cannot be drawn. The flows reported to be entering a country receive more emphasis than flows out of a country. Additionally, the disadvantages of imported goats were mentioned in only a very few reports. Only for Africa was the increased susceptibility to disease of exotic improved breeds addressed in several country reports (Table 1). 238 CURRENT STATUS OF GOAT GENE FLOW

Table 1 summarises the information from Country Reports on goats. Some country specific information can be depicted in Annex 9.1 A 8. On the whole, information on goats is scarce and less comparable between countries than that of other domestic livestock species.

Table 1: Country Reports on gene flow of goats

Breed Extent of gene Main breeds Breed Disadvantages Country report flow in and out of imported replace- of exotic breeds excerpts containing the region ment information In Out Africa medium 0 Boer, Anglo low Susceptibility to Botswana, Burundi, Nubian diseases Ethiopia, Uganda Eastern low 0 Lithuanian low Not named Russian Federation and Blackhead, Southern Alpine Europe Latin none America North medium low Nubian, Alpine, low Not named Canada, United States America Saanen of America Oceania low 0 Anglo Nubian 0 Not named Palau Southern low 0 Saanen, Boer, 0 Not named China, Indonesia, and Jamnapari Japan, Nepal, Eastern Philippines, Sri Lanka Asia Western none and Northern Europe Western none Asia REFERENCES GLOBAL GENE FLOW OF GOATS 239

4 REFERENCES GLOBAL GENE FLOW OF GOATS

Arbiza, S.I. 1983. Evolution of Mohair, Cashmere and Skins. In: Neimann-Sorense, A. and D.E. Tribe (eds.) World animal science. Elsevier, Amsterdam, The Netherlands. Casey, N.H. and W.A. van Niekerk. 1988. The Boer goat. I. Origin, adaptability, performance testing, reproduction and milk production. Small Ruminant Research 1:291-302. Cutrer, G. 1995. Boer goats for beginners. Ranch and Rural Living Magazine 77(2). Devendra, C. and K. Nozawa. 1976. Goats in Southeast Asia - their status and production. Zeitschrift für Tierzüchtung und Züchtungsbiologie 93:101-120. Erasmus, J.A. 2000. Adaptation to various environments and resistance to disease of the Improved Boer goat. Small Ruminant Research 36:179-187. Gall, C. 1996. Goat breeds of the world. CTA, Wageningen, The Netherlands. Gall, C. 2001. Ziegenzucht. Ulmer Verlag, Stuttgart, Germany. Gangyi, X., P. Jiabi, Z. Hongping, and C. Taiyong. 2000. A preliminary report on improvement by using Boer goat in China. 7th International Conference on Goat, France, 15-21 May 2000, 342-345. Glimp, H. 1995. Meat goat production and markeing. Journal of Animal Science 73:291- 295. Horst, P. and S.S. Husain. 1991. Animal genetic resources. In: Panandam, J., S. Sivaraj, T.K. Mukherjee, and P. Horst (eds.) Goat husbandry and breeding in the tropics. Compiled papers presented in an international seminar carried out by German Foundation International for Development (DSE) at the Institute for Advanced Studies, University of Malaya, Kuala Lumpur. 100-113. Künzi, N. and G. Stranzinger. 1993. Allgemeine Tierzucht. 1. Ulmer Verlag, Stuttgart, Germany. Lu, C.D. 2002. Boer goat production: progress and perspective. 2001 Conference on Boer goats in China, 21-24 October 2001, Guizhou. 9-17. Luginbuhl, J.M. 2003. Goat production in North Carolina. IGA Newsletter June 2003. 16- 17. Machen, R.V. 1997. Boer goats - introduction and impact in North America. Journal of Animal Science 75:139. Malan, S.W. 2000. The improved Boer goat. Small Ruminant Research 36:165-170. Mason, I.L. 1981. Wild goats and their domestication. In: Gall, C. (ed.) Goat Production. pp. 35-53. Academic Press, London. Montaldo, H.H. and C. Meza. 2000. Goat genetic resources situation in Mexico. 7th International Conference on Goat, France, 15-21 May 2000, 944-945. Murray, P.J., T. Johnson, E.A. Qualischefski, and J.E. McCosker. 2000. The Boer goat down under - revitalizing Australia’s goat meat industry. 7th International Conference on Goat, 15-21 May 2000, France. 221. 240 REFERENCES GLOBAL GENE FLOW OF GOATS

Nozawa, K. 1983. Domestication and history of goats. In: Neimann-Sorense, A. and D.E. Tribe (eds.) World animal science. Elsevier, Amsterdam, The Netherlands. Pashaa, T.N. and S. Saithanoob. 2000. Goat meat production in South and South East Asia. 7th International Conference on Goat, France, 15-21 May 2000, 623-626. Porter, V. 1996. Goats of the world. Farming Press, Ipswich, UK. Sahlu, T. 2000. The American goat situation. 7th International Conference on Goat, 15-21 May 2000, France. 894-896. Scherf, B. 2000. World Watch List for Domestic Animal Diversity. 3rd Edition. FAO publication. http://dad.fao.org/en/refer/library/wwl/wwl3.pdf (01.06.2005). Sohnrey S.B., and W. Holtz. 2000. Transcervical embryo collection in Boer goats. Small Ruminant Research 36:195-200. Tuncel, E. 1987. World production and utilisation of Mohair. In: Santana, O., A.G. da Silva, and W.C. Foote (eds.) IV. International Conference on Goats, 8-13 March 1987, Brasilia, Brazil. pp. 169-175. EMBRAPA, Brasilia. Yonghong, H. 2001. Utilisation and development of Boer goats in China. 2001 Conference on Boer goats in China, 21-24 October 2001, Guizhou. 21-24.

ANNEX 241

ANNEX

9.4 Global gene flow of cattle

M. Mergenthaler, H. Momm, A. Valle Zárate

242 TABLE OF CONTENTS

TABLE OF CONTENTS

List of tables 243

List of figures 243

Abbreviations 244

1 Introduction 245

2 Historical development of cattle gene flow 245 2.1 Gene flow during domestication, migration and breed formation 245 2.1.1 Early gene flow in the Old World 246 2.1.2 The penetration of the New World 247 2.1.3 Breed formation 248 2.2 Systematic influences on gene flow 249 2.2.1 Levels of gene flow 249 2.2.2 Influence of breeding goals and methods on breed development and spread 250 2.2.3 Influence of technology, disease and animal welfare rule on global mobility and dissemination 251

3 Current status and actual trends of cattle gene flow 252 3.1 Gene flow indicated by selected export and import data 252 3.1.1 Europe 253 3.1.2 North America 261 3.1.3 South America 262 3.1.4 Asia 263 3.1.5 Africa 266 3.1.6 World 267 3.2 Gene flow indicated by changes in the breed composition 269 3.3 Gene flow indicated by the introduction of foreign genetic material into existing breeds 271 3.4 Gene flow indicated by Country Reports 274

4 References global gene flow of cattle 276 LIST OF TABLES 243

LIST OF TABLES

Table 1: Export of purebred breeding heifers from EU countries 1993-2003 254 Table 2: Import of purebred breeding heifers to EU countries 1993-2003 255 Table 3: Simmental breeding cattle exports from Germany and Austria 1998-2002 256 Table 4: Semen import to EU-15 countries and insemination with imported semen in 2003 259 Table 5: Cattle breeding imports into the Philippines 265 Table 6: Total numbers of cattle by breed in Switzerland 269 Table 7: Examples of replacement of local cattle by imported Simmentals 274 Table 8: Country Reports on gene flow of cattle 275

LIST OF FIGURES

Figure 1: Possible migration routes of Bos taurus type cattle from western Asia 246 Figure 2: Major cattle introduction routes into North, Central and South America 247 Figure 3: Semen imports and exports of the Simmental breed for upgrading in beef breeds in 1970 252 Figure 4: Share of semen imports to EU countries by origin in 2003 258 Figure 5: Mean annual imports of semen doses to Germany by partner region for 1998-2002 260 Figure 6: Imported Simmental breeding animals to Brazil by country of origin for 1905-1994 262 Figure 7: Number of inseminations by breed and origin of semen in the dairy sector in Mozambique 1983 267 Figure 8: Number of cattle in Queensland/Australia 271 244 ABBREVIATIONS

ABBREVIATIONS

AD anno Domini AI Artificial insemination BC before Christ BHV1 Bovine Herpesvirus type 1 BSE Bovine Spongiform Encephalopathy CIS Commonwealth of Independent States DNA Deoxyribonucleic acid e.g. for example e.V. Eingetragener Verein (Incorporated society) EFRP Emergency Farm Reconstruction Project (Kosovo) EU European Union EU-15 Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Portugal, Spain, Sweden, The Netherlands, United Kingdom FAO Food and Agriculture Organisation of the United Nations GDR former German Democratic Republic ICAR International Committee for Animal Recording Interbull International Bull Evaluation Service UK United Kingdom US United States USA United States of America USSR former Union of Socialist Soviet Republics

INTRODUCTION 245

1 INTRODUCTION

The global gene flow study of cattle consists of two parts, the historic development and the current status of gene flow. Exemplary for the species Bos taurus, the Holstein and Simmental breeds and for Bos indicus the Brahman and Sahiwal breed have been chosen to highlight typical developments of cattle gene flows. Material from various sources were used. Data was sourced from the Eurostat statistical database, Country Reports on the State of Animal Genetic Resources and from reports and publications from regional experts and breeding organisations with special reference to the breeding organisation of Simmental, Holstein and Zebu breeds. Regarding the historical development, influences of domestication, breed formation, human migration and breeding methods on the diversification of cattle are summarised. The depiction of the current status focuses on selected export and import information and on indicators for the introduction of foreign genetic material characterised for regional clusters.

2 HISTORICAL DEVELOPMENT OF CATTLE GENE FLOW

2.1 Gene flow during domestication, migration and breed formation The world cattle population today accounts for 1,339 million cattle. 1,022 million, about one quarter are found in developed countries, the remaining three quarters in developing countries (FAO, 2004). Bos primigenius is believed to be the monophyletic ancestor of today’s Bos taurus and Bos indicus cattle breeds. The first evidence of domesticated cattle appeared on the Southern Anatolian plateau in Turkey about 6,400 BC, and in Greece and Macedonia around the same time (Payne and Hodges, 1997), after sheep and goats. It is believed that cattle were domesticated independently in at least two distinct centres, one in western Asia (Bos taurus and Bos indicus) and the other in Southeast Asia (Bos (bibos) spp.). In Southeast and East Asia, gene flow from the Bos (bibos) wild species to today’s local cattle breeds probably occurred. Domestic Yak and all buffaloes together with the mentioned Bos and Bos (bibos) species belong to the bovinae sub-family (Payne and Wilson, 1999). It has been suggested that the first domestic cattle in western Asia were the longhorn type of Bos taurus. The shorthorn type appeared about 1,000 years later. There are several theories on the ancestry of shorthorn type cattle, but it is likely that they shared a common ancestor with longhorn cattle and arose through conscious selection. Bos indicus, Zebu cattle are humped. It is unknown whether the Zebu resulted from the domestication of a specialised strain of aurochs or Asian urus (Bos primingenius namadicus) which was already adapted to hot, dry environments, the crossbreeding of domestic and wild cattle, or selection by man in existing domestic herds (Payne and Hodges, 1997). It is assumed that Bos (bibos) spp. were domesticated in two distinct centres: the gaur stemmed from regions of Bhutan, Myanmar and parts of Bangladesh and India, and the 246 HISTORICAL DEVELOPMENT OF CATTLE GENE FLOW banteng were most likely domesticated in the Indonesian islands of Java and Bali (Payne and Hodges, 1997).

2.1.1 Early gene flow in the Old World

During the Neolithic transition, migrating tribes took along their animals and thereby affected gene flow in domestic animal species very early in history. They brought cattle from Western Asia into Central Asia, the Indian subcontinent, Africa and Europe (for a graphical depiction see Figure 1 and Annex 9.1 A 9 to A 11). Starting in the late 18th century but mainly effective during the 19th century, important cattle migrations occurred from the alpine area to all parts of the Austrian-Hungarian empire and the Balkans (Simmental, Brown Swiss, Pinzgauer) and from the German-speaking countries to Russia (including lowland cattle). Migration led to the gradual extinction of wild African and European aurochs. Zebu cattle were generally spread at later stages than taurines, entering the African continent from the primary domestication centres in Arabia and India. In eastern Africa, crosses of Zebu cattle with existing taurine cattle formed the Sanga. From there, they spread throughout Eastern and Southern Africa and eventually to the whole continent. In western Africa, taurine cattle developed a trypanotolerance through natural selection. Zebu cattle arrived in West Africa only in the last 1,400 years through Arabic influences and possess little trypanotolerance. (Epstein and Mason, 1984; Machugh et al., 1994; Machugh, 1996; Payne and Wilson, 1999). A study by Hanotte et al. (2002) based on microsatellite marker analysis of continent-wide samples suggests both indigenous domestication of cattle in Africa and genetic influence from other domestication centres in West Asia and the Indus Valley.

Figure 1: Possible migration routes of Bos taurus type cattle from western Asia

Source: Payne and Hodges (1997) HISTORICAL DEVELOPMENT OF CATTLE GENE FLOW 247

2.1.2 The penetration of the New World

In North America first human-induced gene flow of cattle (together with sheep and goats) took place in 1000 AD by the Norsemen - which did, however, not last. Cattle, sheep and pigs were brought to Latin America by the Spanish and Portuguese colonisers in the 16th century, from where they spread over the whole continent (Figure 2). At the beginning of the 17th century British colonists re-introduced cattle to New England from Great Britain. After initial set-backs they were established permanently. In the course of time local strains developed through natural selection which were adapted to their environments (Alba, 1978; Payne and Hodges, 1997; Martínez et al., 2000).

Figure 2: Major cattle introduction routes into North, Central and South America

Source: Payne and Hodges (1997)

Australian settlers brought cattle to Australia in their efforts to create living conditions similar to their homeland. The first imports took place at the end of the 18th century and consisted mainly of British cattle breeds. They were meant to insure against food deficiencies. Although there was also gene flow of Zebu and Criollo cattle from Latin America, it did not have a pronounced effect at the time (Allen, 2002; Daly, 1981; Payne and Hodges, 1997; Payne and Wilson, 1999; Vercoe, 1989). In the second half of the 20th century British breeds were gradually replaced by cattle with Zebu blood (Annex 9.1 A 12). 248 HISTORICAL DEVELOPMENT OF CATTLE GENE FLOW

These first cattle introduced to Latin America, North America and even Australia had a subsistence function similar to that of cattle transported in migrations from West Asia to Europe and Africa. The Criollo cattle of Latin America result from this. Since the middle of the 19th century, however, the gene flow has become increasingly commercial, first through registered pedigree animals and, since the 1960s, complemented by semen and other biotechnology. In addition, gene flow has occurred during or after acts of war and - more significantly - in transfers through livestock trade. This is testified for centuries by established trading houses, empires, and long trade routes at land and at sea.

2.1.3 Breed formation

Formal breed formation began in the 18th century in Britain with the foundation of breed societies which practiced deliberate selection. British cattle breeders, both in Scotland (Galloway, Aberdeen Angus, Ayrshire) and England (Shorthorn, Hereford, Jersey, Guernsey), pioneered the development of breeds which later significantly influenced cattle breeding worldwide. The early work of the breed societies strongly emphasised formalistic criteria through phenotypic standardisation, systematic inbreeding and restrictive pedigree registration rules. The breed societies promoted closed herdbooks in some British breeds, many generations of upgrading in others and the transfer of these ideas to the whole New and English-speaking world. All this contributed to the depletion of genetic variability. This pedigree breeding concept which appeared to be successful in Britain, spread to the European continent, in particular to the Netherlands, where Black and White as well as Red and White cattle were selected, and to Switzerland, where Brown Swiss and Simmental cattle were bred. The Simmental breed is an example of a dual-purpose breed for beef and milk with global influence. The breed ranges from the dual-purpose type, nowadays found mainly in Central and Eastern Europe and the Balkans, to a specialised beef type found mainly in the New World. For details on the total Simmental population per country in the early 1990s, refer to Annex 9.1 A 13. In addition to the Simmental, three other cattle breeds with global outreach will be referred to occasionally in describing gene flow: the Holstein as a single- purpose dairy breed with the highest potential for milk yields under favourable management and environmental conditions; the Sahiwal as a dual-purpose beef and milk breed adapted to tropical and subtropical climates; and the Brahman as a beef breed with good adaptation to harsh environmental conditions. The Simmental breed might suffice as a specific example of breed formation. The Red spotted cattle of Western Switzerland - which includes the Simmen valley - are assumed to have been introduced by the South German Alemanic tribes in the 5th century (Felius, 1995). Further gene flow can be identified by the introduction of large German cattle in the middle ages (Bo et al., 2001). Thereafter purebreeding was practiced and no remarkable gene flow took place. In the 19th century breed societies were founded and several local strains developed, which finally resulted in the “Simmentaler Fleckvieh” (Felius, 1995). More specific insights into the spread of genes can be illustrated by the spread of the Simmental breed from Switzerland. First exports of the Bernese cattle (early designation of the Simmental breed) to Italy are reported from 1480. During the 18th century a large number of these cattle were exported from Switzerland to Bavaria/Germany, HISTORICAL DEVELOPMENT OF CATTLE GENE FLOW 249

Lombardy/Italy and Eastern France. Rural nobility in Central West Russia first imported Simmental to Russia in 1850. In the second half of the 19th century gene flow to Russia, Eastern Europe, Southern Germany, Eastern France, Northern Italy and Austria increased and reached a peak. Crossing with local cattle strains led to their absorption and the development of several Simmental strains (Felius, 1995). Imports from Russia to Inner Mongolia/China in 1898 established an early influence of the Simmental on the Chinese cattle population (Qiu Huai et al., 1993). In 1957 official imports of Simmental from the Soviet Union followed (Liang, 1988). In Southwest Africa (now Namibia) and South Africa the early imports of Simmental by German settlers led to the successful establishment of the breed in the 19th century (Rusch, 1988). Long-term breed comparisons in Namibia and in Omatjenne, Mara and Vaalharts research stations in Southern Africa in the 1950s proved the Simmental’s viability in ranch conditions and produced a wave of Simmental gene flow in the late 1960s. The success in Southwest Africa led to the spread of the Simmental to Latin America, North America and finally also Oceania. Changes in consumer preference to leaner meat catalysed and favoured the Simmental’s utilisation in comparison to traditional European beef breeds (Neser et al., 2002; Rocha et al., 1987; Sonn, 1985). Other large-sized continental cattle breeds also experienced increased gene flow during this time. Simmentals are reported to have been present in the 19th century in Illinois/USA (American Simmental Association, 2004). The first herd of Simmental cattle in the USA was set up in 1886 in Texas but did not lead to wider distribution (Felius, 1995). It was not until the 1960s that the Simmental became fully established in America through imports of semen from Canada (from a French bull out of a Swiss bloodline). The first gene flow to South America is reported for Brazil in 1905 and for Argentina in 1922, but this gene flow did not lead to sustained and lasting establishment of the breed. Considerable numbers of Simmentals imported to Argentina in 1967 were lost as the animals were not pre-immunised, but eventually the Simmental breed was established and spread (Sonn, 1985). In all these regions the Simmental has been used for beef production and has been developed to this end.

2.2 Systematic influences on gene flow

2.2.1 Levels of gene flow

Gene flow may occur at different levels and in different forms. At the first level, the breed is spread through the transfer of breeding females for milk or beef production. At the second level, stud herds are established, primarily to produce bulls for natural service. At the third level, genetic materials are provided for AI stud breeding. The latter includes both the transfer of animals, semen and embryos selected on the basis of contract matings (in exchange programmes of test bull semen and after completing the progeny test) and comparison of the breeding values of bulls at global level through Interbull. The effects of the various levels of gene flow differ. For example, the impact of exporting large numbers of breeding heifers primarily for commercial use as dairy cows is relatively small compared to that of establishing nucleus herds for stud breeding, importing genetic materials through an existing AI network, or participating in an AI breeding programme, provided there is an operational distribution network in the importing country. Although 250 HISTORICAL DEVELOPMENT OF CATTLE GENE FLOW the trade in breeding animals is the simplest form of gene flow, semen and embryos are powerful in transgressing disease barriers with significant multiplier potential. The French example of upgrading their Black and White cattle to Holstein mainly through the import of very valuable embryos is a good example of strong multiplier effects. The basis for this development was the import of less than 1,000 embryos (Meyn, 2005). On the other hand, the widely discussed export of Simmental heifers from Germany (and Austria) to Kosovo may be regarded as a minor event in terms of gene flow.

2.2.2 Influence of breeding goals and methods on breed development and spread

Domestication of wild species generally leads to higher variability within the species, which in turn allows new selection possibilities. Moreover, moving animals to new habitats prompts natural adaptation. These two factors cannot be clearly distinguished in early breed development. Natural and artificial selection lead to the development of many local breeds, that are characterised by high variability which then permits further selection with different breeding goals. Early breed selection focused on conformation with limited advances in productivity. Only later did selection based on performance and genetic evaluation lead to more rapid progress. Many different scenarios exist in which the selection of cattle may occur: • according to utility or market requirements for dual- or multi-purposes, dairy or beef; • according to climate for the temperate, hot-dry or hot-humid zones using the different breed groups of cattle: taurine, Zebu or other cattle; • through pure- or crossbreeding. Dairy cattle breeding in the temperate zone is heavily influenced by the overwhelming market power and outstanding yield of the Holstein. This endangers the existence of other temperate zone dairy and dual-purpose breeds. Some examples of breeds competing with the Holstein are the Simmental, the Montbeliard (as a Simmental subpopulation), the Brown Swiss and the Swedish Red and White/Ayrshire. In Holstein breeding a few top bulls and top lines dominate, due to extremely high selection intensities and internationally-connected breeding populations. This reduces the effective population size and leads to both inbreeding and a higher frequency of recessive gene defects. This way of breeding is most efficient in achieving genetic progress in single traits and building up high-yielding specialised breeds. On the other hand, it increasingly displaces other breeds, impacting gene flow and reducing genetic resources. Nonetheless, appropriate use of progeny testing (including sire x daughter matings) and DNA analysis for genetic abnormalities can be powerful tools for detecting and eliminating undesirable genes. One important advantage of long-term semen storage is that large quantities of semen are available at the time the results of progeny testing become available. Purebred Holstein are also expanding into the dry tropics and subtropics - for example in the subtropical part of the USA, in Israel and Saudi Arabia - and actually achieving their highest yields there. Because of high fodder production costs, high yields with relatively high producer milk prices are required for a breed to be profitable. However, when cattle producers need multi-purpose cattle or if product prices do not justify costly inputs, other breeds and breeding methods have to fall into place. Here, the Sahiwal (Zebu) and its crosses with temperate zone dairy and dual-purpose breeds are possibilities. Not only the HISTORICAL DEVELOPMENT OF CATTLE GENE FLOW 251

Holstein, but also in some situations the pure Simmental, Brown Swiss, Jersey or other breeds may be of interest. For the humid tropics - mainly in Latin America, but also in the Philippines - efforts to introduce Holstein cattle have generally failed. Productive dairy cattle for this eco-climate require some taurine blood with a high frequency of alleles determining milk yield combined with a zebuine breed for adaptability. Beef breeds are less endangered by absorption through other breeds because reproduction is mainly through natural service and there is not yet sufficient knowledge about the comparative efficiency of breeds kept on natural pasture. There are many breeds for diverse purposes: B. taurus breeds from Britain (early maturing but moderate in size) and from the European continent (large size but late maturing) for the temperate zone; many B. taurus x B. indicus crosses for the hotter areas (Santa Gertrudis, Brangus, Beefmaster, Droughtmaster, Bonsmara and other crossbreds); Zebu breeds (Brahman, Gir, Guzera, Nelore, Boran and others); Sanga breeds (Africander and Tuli); and other indigenous breeds.

2.2.3 Influence of technology, disease and animal welfare rule on global mobility and dissemination

There are a number of major restrictions to the international gene flow by means of live animals: • High costs prevents trade over long distances unless the animal is extremely valuable. For breeding cattle there are in fact three zonal markets in the world based on the main exporters: Europe, North America and Oceania. There is little overlap between these export zones. • Increasingly strict disease control in international trade has created barriers to the movement of live animals, in the form of complete bans (e.g. out of Africa), veterinary checks and quarantines, and - increasingly infrequently - vaccinations. In the EU, for example, internal barriers for the movement of breeding stock are very strict due to threats of contamination with Bovine Herpesvirus type 1 (BHV1). Other restrictive diseases are Bluetongue, Scrapie, and many other viral diseases. Severe consequences followed the outbreak of Bovine Spongiform Encephalopathy (BSE). A temporary ban on semen and embryo trade as well as a permanent ban on import of live animals from certain countries were put into place. • Stricter regulation of animal welfare during transport limits the distance live animals can be moved. • Where high-yielding taurine breeds are crossbred with adapted cattle in a tropical climate, live animals of the exotic breed may struggle to survive in the environment. Artificial insemination (AI) was developed first in the former USSR by Ivanov, then in Kenya by Anderson in 1935, and then in the UK. While AI in cattle came into regular use in the USA and UK in the 1940s, it only became commonplace in most countries after the second world war. In some developing countries, the development of AI programmes are still in experimental stages (Chupin and Schuh, 1993, Chupin and Thibier, 1995). The international gene flow in cattle increased substantially after the technique of deep-freezing 252 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW bovine semen was developed in the 1960s, in part because transport costs were reduced and animal welfare and disease control regulations could be more easily complied with. A global inventory of cattle semen imports and exports in 1991 covering 31 developed countries (Chupin and Thibier, 1995) gives some insight into the impact of biotechnology and global mobility on the dissemination of genetic material. 10 countries exported more semen than they imported (Belgium, Canada, Denmark, Estonia, Finland, France, Netherlands, New Zealand, Norway, Slovenia). Altogether more than 4.3 million doses from dairy breeds, 900,000 doses from beef breeds and 760,000 doses from dual-purpose breeds were exported to other regions of the world in 1991. Presumably the USA are the largest net-exporter of semen, but figures on semen import have not been provided. Most of the surplus semen was of the Holstein breed and imported by developing countries. Following the development of techniques to deep-freeze bovine embryos in the 1980s, this technology has also become an important vehicle for gene flow, especially because the risks of disease contamination are minimised through trypsin washing. The Simmental is an historic example of the inter-relation between semen imports and exports and desired global mobility. Figure 3 illustrates the direction of semen trade for purposes of upgrading beef breeds in 1970 and indicates 4 exporting, 3 importing and exporting, 5 exclusively importing countries or regions. This example shows the ingenuity of exporters in overcoming veterinary barriers. Currently, the EU may export semen and embryos directly to all of these countries.

Figure 3: Semen imports and exports of the Simmental breed for upgrading in beef breeds in 1970

Sweden Canada

USA Germany Australia France Austria UK and New Zealand Switzerland

South Africa South America Source: adapted from Künzi and Stranzinger (1993)

3 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW

3.1 Gene flow indicated by selected export and import data This chapter compiles information on selected cattle exports and imports in Europe, North America, South America, Asia and the Middle East. For Europe, a statistical database was accessible from Eurostat (2005) which tabulates information on the exchange of breeding stock from the EU-15 countries. Information on breeds is not included in the database. Currently, the dairy cow population in the EU-15 countries is the most important in the CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW 253 temperate zone, attracting 5.5 million doses of semen from the USA, Canada, New Zealand and Australia and exporting more than 3.6 million doses to third countries in 2004. For other regions, information was extracted from publications and from communication with regional experts. The result is, in general, only illustratory in nature and no claim is made of its completeness. Selected information of breeding animal transfer at breed level was added from various breeding organisations, with special reference to the Simmental, Holstein, and Zebu breeds. Based on these sources, emphasis was put on analyses of the direction of animal transfer and of the observed fluctuations in the export and import of breeding cattle.

3.1.1 Europe

Live animals Annual fluctuations in the exchange of breeding animals are influenced by special export programmes, trade restrictions due to sudden disease outbreaks or long-term veterinary restrictions, and political conflicts. The influence can be positive or negative depending on the specific situation. In general, animal health regulations or measures tend to restrict gene flow, however, they can also promote the exchange of breeding animals. For example, in Germany after 1950, eradication programmes of bovine tuberculosis, brucellosis and leucosis created the demand for disease-free, productive animals. As healthy cattle were not available locally, foreign stock of both the same and foreign breeds were imported. As a consequence Black-and-White and Jersey cattle appeared in the Allgäu region of Germany, where formerly Braunvieh cattle had dominated. Table 1 presents data on the export of breeding heifers from EU countries over eleven years (1993 to 2003), allowing consideration of both year-to-year fluctuations and of the overall direction of gene flow through breeding heifer trade. On average, EU countries exported more than 150,000 breeding heifers annually between 1993 and 2003. Annual export increased from almost 125,000 heifers in 1993 to over 180,000 in 1995. However, heifer exports subsequently declined to around 100,000 in 2001 following the BSE crisis (Table 1). 254 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW

Table 1: Export of purebred breeding heifers from EU countries 1993-2003

Period Extra-EU Intra-EU Total 1993 67,909 55,834 123,743 1994 80,693 78,590 159,283 1995 124,135 57,856 181,991 1996 133,451 56,050 189,501 1997 84,436 77,551 161,987 1998 98,794 73,053 171,847 1999 96,338 83,824 180,162 2000 76,941 74,400 151,341 2001 11,423 87,890 99,313 2002 30,593 102,354 132,947 2003 35,569 65,453 101,022 Roughly half the breeding heifers exported from EU countries remained within EU-borders and half left the EU. Of the latter, the largest share - about 64% - was exported to the West Asia/Maghreb region. The most important trade partners were in Northern Africa. Morocco received 18% of exports leaving the EU, Algeria received 9%, and Tunisia, Egypt, Libya and Sudan together received 7.8%. In West Asia, Turkey received 20% and Lebanon and Jordan, both received 2-3% of exports. The large Turkish share resulted from a special credit programme for the export of German breeding cattle to Turkey which ran from 1988 to 1996 and financed the purchase of 177,000 heifers. 45,000 heifers were exported during its last year alone. Since then, however, virtually no exports have gone to Turkey (Annex 9.1 A 14). The remaining 3.8% of EU exports to West Asia/Maghreb went to 11 countries in West Asia. 34% of heifer exports leaving the EU were received by non-EU European countries. About 19% went to Eastern European countries, most notably Poland, the Czech Republic, Russia and Ukraine, while 14% went to South Eastern European countries. The Balkan countries (Croatia, Bosnia-Herzegovina, Serbia and Montenegro) predominated. Switzerland accounted for 0.7% of the heifer exports leaving the EU while Northern-European countries constituted 0.5% (mainly the Baltic states). Heifer exports from the EU to other parts of the world averaged 1.8% annually. Of these, Brazil and the Philippines were the only two countries receiving more than 0.1% of extra-EU exports. For more details refer to Annexes A 15 and A 16. From 1993 to 2003, the EU was a net-exporter of purebred breeding heifers. Exports exceeded imports by almost 40,000 per year (import-export-ratio: 0.7). However, the absolute number of trades as well as the import-export-ratios of individual countries showed considerable variation. The main net-exporting countries were Germany, Denmark, the Netherlands, Sweden and France, Austria and Finland. However, since genetic exchange from Sweden and Finland is restricted mainly to Norway and Denmark, they can not be considered true exporters. Additionally, the UK was an importer of breeding heifers during the years of recovery from BSE. CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW 255

The two main importing countries are Spain and Italy. Their import-export-ratios were 27 and 22 respectively. Other net-importers were Portugal, Greece, the UK, Ireland and Belgium-Luxembourg. EU imports from the Czech Republic, Slovakia and Hungary were mainly cheap beef and dairy cattle for herd build-up, e.g., in the former GDR. In Italy, Spain, Portugal and Greece, there was a tendency to import replacement breeding heifers from the north (France, Germany, the Netherlands, Belgium, Austria and Denmark), where rearing costs are lower. A decline of imports was caused by the BSE crisis. For a comparison of absolute numbers refer to Annex 9.1 A 17. In Table 2, import data for EU countries over the 1993 to 2003 time period are presented. On average, EU countries imported more than 111,000 breeding heifers annually from 1993 to 2003. Imports to EU countries ranged from 80,000 to 130,000 head of cattle, without a clear trend. A peak of more than 200,000 imports in 1994 resulted from a massive Spanish import of almost 70,000 French breeding heifers. Italy has become the largest importer within the EU, followed by Spain, Greece and Portugal.

Table 2: Import of purebred breeding heifers to EU countries 1993-2003

Period Extra-EU Intra-EU Total 1993 65,186 66,120 131,306 1994 69,687 131,599 201,286 1995 34,394 83,346 117,740 1996 10,781 82,369 93,150 1997 6,266 104,226 110,492 1998 4,351 98,046 102,397 1999 3,189 89,317 92,506 2000 3,008 114,052 117,060 2001 1,267 71,592 72,859 2002 902 102,776 103,678 2003 1,129 79,139 80,268 In the period studied, only 13% of the imports to EU countries came from extra-EU countries. The remaining 87% are intra-community trade and cannot be considered real imports. The increase in the percentage of EU “imports” originating within the EU - from 67% in 1993 to 99% in 2003 - may be explained by the accession of exporting countries of breeding cattle (Austria, Sweden, Finland) and by the BSE scare. Imports during the early 1990s were mainly intended for initial herd build-up, as was already discussed. Of the imports from extra-EU countries the majority (88%) were from Eastern Europe. The most important trade partners were the Czech Republic (60%), Hungary (22%), Slovakia (2.8%) and Poland (2.8%); all are EU-member states today. 8.1% of imports came from Switzerland and imports from six other European countries combined to make up slightly more than 1%. Canada dominated imports from North America with 3.3%. Marginal shares of imports to the EU came from the USA, Algeria, Morocco, Hong Kong and other unspecified countries, which together made less than 1% contribution. Imports from the USA were restricted due to Bluetongue, Bovine Leucosis and BHV1. No imports were 256 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW reported from other parts of the world. For absolute numbers, refer to Annexes A 18 and A 19. Table 3 presents data summarising exports of Simmental cattle from Austria and Germany over a five year period (1998 to 2002). On average almost 7,500 breeding animals per annum were exported from Germany and more than 10,000 were exported from Austria. After 2000, the German export figures were more strongly affected by the BSE crisis than the Austrian ones. Of the breeding cattle exported from Germany, 85% were heifers, 5 % cows, 5% calves, 4% maiden heifers and 0.5 % bulls. In Austria the exports were 73% pregnant heifers, 22% cows, 4% heifers, 1% calves and 0.5% bulls. Exports to European countries were dominant, 88% and 98% of German and Austrian Simmental exports, respectively. 67% of German and 46% of Austrian exports went to East and Southeast European countries, predominantly Bosnia, followed by Croatia and Kosovo. Bosnia and Kosovo were important buyers of cattle because of donor-funded reconstruction programmes.

Table 3: Simmental breeding cattle exports from Germany and Austria 1998-2002

Year Austria1 Germany2 1998 9,850 13,174 1999 13,185 9,585 2000 10,439 8,411 2001 8,728 3,190 2002 8,026 3,072

1 data from Arbeitsgemeinschaft Österreichischer Fleckviehzüchter (2004), 2 data from Arbeitsgemeinschaft Deutscher Rinderzüchter e.V. (2002)

21% of German Simmental exports went to EU countries, while more than 50% of the Austrian Simmentals remained in the EU. Italy was the biggest customer, followed by Spain (for Germany) and Germany (for Austria). The main destination for exports of Simmental breeding animals outside Europe was West-Asia/Maghreb. 12% of German and 2% of Austrian exports went to this region. The dominant importers were Morocco, Algeria and Egypt (for Germany) and Algeria (for Austria). Until 1996, Turkey had overriding importance. Between 1998 and 2002, Austria had no other exports while 0.2% of Germany exports went to Colombia and an even smaller share of breeding animals went to Australia. Considering a longer time span, Averdunk et al. (2001) report export booms from Germany to Italy and Yugoslavia in the 1970s and to countries of the Maghreb region in the 1980s. In the 1970s exports went to North American and Latin American countries, to New Zealand and Australia, and to countries in Northern Europe (Figure 3). The foundations of Simmental populations in these regions were based on the introduction of relatively few animals or imports of semen. As the Simmental breed became established and expanded, and breeding organisations were founded (e.g., the World Simmental Federation in 1974), exports to these countries have been limited to a few animals of high breeding value (Grupp, 2003; Jakopovic et al., 1995; Rusch, 1988). Simmentals were exchanged between the new breeding associations and exported to countries in which they CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW 257 were not yet established (Felius, 1995). Political and economic developments in African and South American countries at the end of the 1970s are cited as contributing factors for dropping exports to these regions (Sonn, 1985), but the apparently low replacement requirements of beef breeding herds may also play a role. Simmental exports from Germany to other regions of the world have been less pronounced. For example, in 1976 about 3,000 Simmentals were exported from Germany to Heilonjiang Province, Manchuria (Meyn, 2005). Between 1979 and 1984 more imports followed from Austria, Switzerland and France (Liang, 1988). Yearly exports of 500 to 1,500 German Gelbvieh and Simmental breeding animals by a German Cattle trading company to ranching regions of Kenya, Uganda, Zambia, Angola, Mozambique and South Africa (Sonn, 1984) are reported from 1965 to 1975, with Simmentals always outnumbering Gelbviehs. In all these regions, Simmentals were developed as a dual-purpose breed with more or less equal emphasis on meat and milk. For more detailed information on the total exports over time see Annex 9.1 A 20. Annex 9.1 A 21 and A 22 compile information on environments to which Simmentals have been introduced and their crosses with local breeds in the second half of the last century. In the 1970s, Holstein heifers were imported to Yugoslavia and a genetic improvement programme was started. Simmental, Black and White, Red and White, and other domestic cattle breeds were upgraded with Holstein up to Holstein purebreds (Medic et al., 2002). From June 2000 to June 2003, the Kosovo Emergency Farm Reconstruction Project (EFRP) was implemented to reconstruct agricultural production in areas damaged during the Balkan conflict (Cossee, 2003). Although not all imported animals originated from Germany the impact on export numbers was significant because of the import ban by many other countries because of BSE. The share of Simmental exports to Kosovo rose from 1% in 2000 to 9% in 2001 and 36% in 2002 (Arbeitsgemeinschaft Süddeutscher Rinderzucht und Besamungsstationen e.V., 2002). According to the World Bank (2003), breeding cattle imported under the project included 4,399 heifers and 92 bulls. By the end of 2003 the herd had grown to 9,000 head. The effect these transfers had on gene flow is unclear, since animals were primarily used to build up milking rather than breeding herds. Absolute figures of German Simmental breeding animal exports to other Balkan countries can be seen in Annex 9.1 A 23. The export ratio of breeding animals gives an indication of the relative importance of gene flow within a country and gene flow that leaves the country. Between 1998 and 2002 the export ratio of Simmental breeding animals in Germany was 30%. Though inland sales were more stable than exports in this period, there is some indication that years in which export sales were high, inland sales decreased (Arbeitsgemeinschaft Süddeutscher Rinderzucht und Besamungsstationen e.V., 2002). The exports of the Swiss Simmental Breeding Association serve as another example of fluctuations in trade caused by disease. While the Association exported almost 6,000 breeding animals to eight different countries in 1996, its export business collapsed entirely in the following two years due the outbreak of Bovine Spongiform Encephalopathy (BSE) and the associated restrictions Switzerland had to face in trading live animals. Only in 1999 did their exports slowly start to recover (Schweizerischer Fleckviehzuchtverband, 2000); from 2003 there were announcements of revived exports (Agro-News Aktuelles aus der Landwirtschaft, 2003). 258 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW

From the mid 1970s to the outbreak of BSE in Switzerland around 1991, the government paid generous export premiums to support breeding cattle exports in competition with the export premiums of the EU. Together, Switzerland and Austria were able to supply all the breeding cattle required by Italy. Export premiums were also important instruments for breeding cattle exports of the EU to third countries, but their relative values became smaller and smaller over the years and they are now expiring with the agricultural reform 2003 of the EU. Croatia was an important importer of breeding cattle in the 1970s, then became a net- exporter and remained a net-exporter until the Balkan war in the early 1990s. As agricultural assets were lost in the war, Croatia depended on imports of food products in the post-war period. For reconstruction purposes, foreign breeding animals were imported and Croatia again became a net-importer of live breeding animals (Haluska, 2002). Semen According to Eurostat, the internal EU export trade in semen was 3.7 million doses at a trade value of € 23.4 million in 2003. Almost 3 million doses were exported to third countries at an export value of € 15.6 million. Notably, not all the quantities are reported in the database, but the export values are all given. According to these statistics, export shares in doses of semen are: other Western Europe and accession countries 37%, Latin America 21%, CIS, Balkan countries, Turkey and North Africa 17%, USA/Canada 13%, Australia/New Zealand 7%, Asia except CIS and Turkey 3% and Africa except North Africa 2%. For more details on the absolute numbers of semen exports see Annexes A 24, A 25 and A 26. While North America is a significant exporter of semen, its role as an importing region is minor, according to Chupin and Thibier (1995). Among developed countries, the EU is the primary importer of cattle semen, followed by non-EU Western European countries, other developed countries (Australia, New Zealand, South Africa), and Eastern European countries. This finding is confirmed on the basis of imported semen value (Hemme, 1995). In 2003, EU countries imported about 6.8 million semen doses. Most of the imports originated within the EU and the majority of the non-EU imports came from the USA and Canada. A very small percentage of imports originated in Australia/New Zealand and other countries (Figure 4). For more details see Annexes A 27 and A 28. Some imports resulted from joint testing programmes with other countries (e.g. Czech Republic, Slovenia or Switzerland) and returning semen of test bulls to their countries of origin (e.g., Hungary, Czech Republic, Poland or Romania).

Figure 4: Share of semen imports to EU countries by origin in 2003

Number of doses Share in % intra-EU 2,926,437 39 USA 2,476,008 33 Canada 1,759,377 23 Others 366,551 5 Total 7,528,373 100

Source: Eurostat (2004) CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW 259

The Netherlands, Germany, and the UK were the largest semen importers in the EU. AI coverage is an indicator of a country’s accessibility to exogenous gene flow. The ratio of the number of doses of semen imported to the number of cows or first inseminations carried out (Table 4) provides a measure of the accessibility, awareness, and penetration of foreign semen in a given country. This relation indicates the awareness, interest and willingness of the cattle producers in the country to import semen. For a comparison of imports and exports in absolute numbers including a global view of the early 1990s see Annexes A 29, A 30 and A 31.

Table 4: Semen import to EU-15 countries and insemination with imported semen in 2003

Country Imported semen doses Inseminations with imported semen (%) Netherlands 1,527,180 38 Germany 1,449,358 53 UK 1,261,812 18 Spain 846,744 42 France 593,781 11 Belgium 566,127 30 Austria 376,051 48 Italy 262,532 22 Portugal 195,494 19 Sweden 169,830 26 Finland 115,499 5 Denmark 77,131 14 Luxembourg 60,713 36 Greece 28,987 6 Ireland 1 78 EU-15 7,531,240 23

Source: Eurostat (2004)

For a detailed trend analysis German semen imports and exports were analysed. Between 1998 and 2002, on average 290,000 semen doses of the Simmental breed and more than 677,000 doses of the Holstein breed were exported from Germany per annum. The percentage of all semen exports which were Simmental in this period was relatively constant at 23% (max: 26%; min 18%). 55% of semen exports were Holstein. Refer to Annexes A 24 and A 25 for data on German import and export rates and their development over time. The notable depression of German semen exports from 1991 to 1994 resulted from re-unification. The impact of the BSE crisis can also be seen in the data for 2000 and 2001, when it was not yet established that BSE cannot be transmitted by semen. Germany has a long history of semen imports from North America. Imports increased in 1994 to 1995 due to the entry of Austria into the EU. 260 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW

On average, more than 26,000 doses of Simmental and 573,000 doses of Holstein semen were imported into Germany annually. The percentage of imported Simmental doses from first inseminations with Simmental is 1.4%. It is not specified where the imported semen doses came from. Simmental semen comprised 4.7% of all semen imported between 1998 to 2002 on average, with considerable variation (max: 6.1%; min: 0.6%). For the Holstein breed, this percentage was 86%. A comparison of Simmental and Holstein over this period is not useful, however. Large-scale gene flow of the Holstein breed from North America to Europe was still ongoing, but there was no similar movement of the Simmental breed because the Germans were the top breeders of the Simmental and did not consider it worthwhile to buy semen from other countries. An average of 662,346 cattle semen doses were imported to Germany between 1998 to 2002 (all breeds). More than half the imports originated from North America, while 34% came from EU countries, 18% came from Eastern Europe, 0.9% came from the remaining West European countries (effect of BSE crisis, especially Switzerland), and 0.3% came from Australia and New Zealand. No semen was imported from other regions of the world in the years studied. German AI organisations frequently keep test or lay-off bulls in various Central and East European countries and re-import semen, which explains the large percentage of imports originating in Central and East European countries. For absolute numbers of mean imports see Figure 5 and for more details on their development over time see Annex 9.1 A 32.

Figure 5: Mean annual imports of semen doses to Germany by partner region for 1998- 2002

North-America 417,562

EU 227,948

Eastern Europe 10,871

remaining W-Europe 4,338

Australia and New 1,627 Zealand

0 250,000 500,000 number of semen doses

Source: Arbeitsgemeinschaft Deutscher Rinderzüchter e.V. (2002) Recent trends in semen and embryo exchange for internationally-integrated breeding work are illustrated by the following two examples. Between 1998 and 2002, Germany exported an average of 158 Simmental embryos annually (Arbeitsgemeinschaft Süddeutscher Rinderzucht und Besamungsstationen e.V., 2002). From 2001 onwards, German Simmental embryos of top-mating pedigree were implanted in Canadian recipient cows and female progeny were used in North American dairy farms. Semen was returned to Germany, where male progeny were used as test sires. These cattle served as insurance in case of epidemics in Europe (Grupp, 2003). CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW 261

An intensive exchange of Simmental semen in Europe is indicated by a common test bull programme in Austria, Switzerland, Germany, France, Italy, Slovenia, Czech Republic and Slovakia (Dodenhoff and Egger-Danner, 2004).

3.1.2 North America

In 1973, eight plane loads of breeding cattle were exported from the USA to Honduras. The majority of these animals were of the Brahman breed, though Santa Gertrudis and were also included. Annex 9.1 A 33 shows us Brahman exports by destination over a 5-year period. In the previous year, Holstein breeding animals had been imported (McDonald, 1973). Beefmaster genetic material, a synthetic breed with major contribution from Brahman cattle, was exported from the USA to Latin America and Australia in the 1970s and 1980s (Schuhmacher, 1983). Similarly, Brangus genetic material (a synthetic breed of Brahman and Angus) was exported from the USA from the 1970s onwards and led to worldwide distribution (Holbert, 1983). The spread of the Holstein is one of the most significant gene flow events in history. The Holstein was bred as a single-purpose dairy breed from the 1950s onwards. The large size of the cow, its high milk yield, and American salesmanship combined to drive extensive exports from the USA which made the Holstein the single most important dairy breed in the world. The genetic spread occurred primarily via export of semen and embryos to the developed world, with France being the main importer of embryos. However, between the 1950s and the early 1990s an estimated 3.9 million cattle of dairy breeds were imported from developed countries by developing countries, 2 million in Latin America and 1.9 million in Africa and Asia. In the early 1990s, there were nucleus herds of high-yielding dairy breeds in 52% of African, 71% of Asian and 100% of Latin American countries. From the 1960s onwards, the Holstein breed was the predominant breed among all imported dairy breeds (McDowell, 1992). Beginning in the 1950s, dairy cattle breeders in Europe became interested in the highly productive North American dairy breeds, in particular the Holstein-Friesian. As grain prices declined, feeding even large quantities of concentrates became economical. Breeders looked for large-framed dairy cattle which had been successfully bred over many decades solely for milk yield, disregarding fat test, conformation and beef characteristics with lactation yields almost twice that of European dual-purpose breeds. In the course of a few years almost all AI stations and breeders associations used genetic resources from the USA and Canada, either in the form of live imported sires and heifers or via import of semen and embryos. In many breed associations this massive gene transfer resulted in almost complete upgrading. Standards for the increasingly observed type appraisal in North America were adopted, though they were sometimes modified to emphasise the old dual- purpose type to some extent. The 4% fat test breeding goal was also conserved. Between 1985 and 1998, domestic sales of semen in the USA declined while exports further increased, raising the export share from 16.5% (2,524,000 doses dairy) to more than 40% (8,789,700 doses dairy, 848,300 doses beef and 18,370 embryos) in 1998 (Monke, 1999). 262 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW

3.1.3 South America

The establishment of the Simmental breed in Brazil was dominated by imports of genetic material from Germany. From a total of 3,001 live breeding animals imported between 1905 and 1994, 66% originated in Germany. Smaller percentages came from the other traditional Simmental-keeping countries: 6% from Switzerland, 5% from Austria, and 1% from France. The remaining 22% arrived from countries where the Simmental became widely established only in the second half of the 20th century, with the largest share (11%) coming from Argentina where the distribution of Simmental in South America started (Stange, 1988). Smaller shares originated in Canada (5%), Uruguay (3%), the USA (2%), and South Africa (0.3%). For absolute numbers see Figure 6 (Fraga, 2004).

Figure 6: Imported Simmental breeding animals to Brazil by country of origin for 1905- 1994

Germany 1.978

Argentine 326

Switzerland 195

Canada 161

Austria 141

Uruguay 91

USA 74

France 25

South Afrcica 10

0 500 1.000 1.500 2.000 2.500 number of animals

Source: own elaboration, data from Fraga (2004)

Between 1986 and 1993, 1,835 Simmental embryos were imported to Brazil. 60% of these imports were Canadian, 26% were German, and smaller shares came from Italy, Argentina, Switzerland, and the USA (Annex 9.1 A 34). 166,065 Simmental semen doses were imported to Brazil between 1972 to 1993. 47% of the semen came from Germany, 27% from the USA, 9% from Canada, 9% from Italy, 6% from Switzerland, 0.9% from Argentina and 0.7% from Austria (Annex 9.1 A 35). The registered cattle population with Simmental blood grew to more than 370,000 animals in 2003, becoming one of the most important cattle breeds in Brazil (Fraga, 2004). Yearly sales of semen doses of the Simmental breed in Brazil were 273,000 in 2002, which corresponds to 3.9% of all semen doses sold (Country report of Brazil, 2003). For the total number of semen doses sold per breed, see Annex 9.1 A 36. Simmental genetic material - both live animals and semen - were introduced to Colombia in the 1970s through collaboration of the Colombian and German governments. The first imported Simmental bulls arrived from Switzerland in 1968, but they did not immediately CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW 263 have a significant influence. Later, however, they led to a gradual spread of the Simmental breed throughout the country. The Simmental was used as a dual-purpose breed in regions with favourable climactic conditions and in crossings with more robust breeds like the Brahman elsewhere. In the mid 1980s, import restrictions on live animals were eased, leading to massive imports of different breeds from Europe and North America (Asociación Colombiana de Criadores de Ganado Simmental y sus cruces-AsoSimmental, 2004; Orbita, 2004 and SOPEXA, 2004). Though Holstein and Zebu cattle are also of considerable significance in South America, information on these breeds was not available. In 1964, a nucleus herd of purebred Central American Criollos was exported from Nicaragua to Mexico. Bulls emerging from these herds were later used in AI of local cattle or exotic breeds. Selection in a Criollo herd in Costa Rica from the 1950s led to the distribution of genetic material from this herd to several Latin American countries, namely Mexico, the Dominican Republic, and Bolivia. The most widespread distribution was reported in areas where milk producers were already organised and established institutions could be used (Alba, 1978; Felius, 1995; Israeli Cattle Breeder’s Association, 2004). The Brahman and the Santa Gertrudis breeds had the most influence on beef cattle breeding in the tropical and subtropical zones of the New World. They were also imported into several African countries, but were found to have shortcomings as subsistence milk suppliers for pastoral herdsmen. The main trading partner of the American Brahman Breeders changed two times during the six years from 1977 to 1982. While Venezuela was the main partner country in 1977, it played a marginal role only three years later, while Mexico became more significant. In the interim, Colombia was the main partner country of American Brahman exports (Cowert, 1983). From 1983 to 1988, 120,000 Holstein animals were imported into Venezuela from North America (Vaccaro, 1990). Imports and exports of other South American countries could not be quantified.

3.1.4 Asia

Asia contains two distinct breeding zones, a temperate zone and a tropical and subtropical zone. In the temperate zone, there is an emphasis on dairy cattle but also some beef. Countries in the temperate zone include the former Soviet Union: Siberia (Russia), Kazakhstan, Uzbekistan, Kyrgyzstan, Turkmenistan, Tajikistan. These countries have a history of continental dual-purpose breeds, Holsteins, and also the “Kazakhian Whitehead” for beef production which is in fact an upgrade of the Hereford breed. In Japan, Korea and China, which are also in the temperate zone, there is an emphasis on Holsteins but also Simmentals (Manchuria). Turkey, Syria, Lebanon, Jordan, Iraq, Iran, Saudi Arabia, Kuwait, United Arab Emirates are all dependent on the import of live animals from Europe. Israel, Saudi Arabia, Iraq and Iran also occasionally import cattle from the USA. In Iran, breeding cattle were first imported in 1940 by the University of Tehran. They originated from France and included the Tarantaise, Tachetee, Brown Swiss and Montbéliarde breeds. Gradual but steady import activity is reported from 1951 onwards. Purebred dairy cows (Brown Swiss, Jersey, Danish Red, Simmental and Holstein breeds) were imported first from Europe and later from North America. While government bodies dominated cattle import in the 1950s, private sector import became more significant with 264 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW time. Imports of live breeding animals ceased in 1980. 95% of the 750,000 purebred cattle kept around large cities in Iran are Holstein (Schahidi et al., 2001). From 1980 onwards, imports of breeding animals into Iran were prohibited in order to encourage purebreeding within the country. From 1983, progeny tests in purebred cows were performed. The first AI station had already been set up by the government in 1950. By the end of the 1990s, there were seven stations. From 1991 to 1997, the share of semen imports decreased from 28% to 5.7%. This is partly explained by an increase in the genetic potential of Iranian bulls, but the exchange rate is also an influential factor. For a long time both cattle imports and the production of semen were limited to governmental bodies. In the late 1990s, political stability increased and the private sector was given more freedom. Private companies began to import semen from the USA, Canada, Holland, and Germany to develop the potential for a dairy industry based on temperate zone breeds (Schahidi et al., 2001). From 1942 onwards, the genetic improvement policy in Iraq called for crossbreeding of local cattle with Holsteins. In 1962, AI was introduced. The number of inseminations increased from 648 in 1962 to more than 132,000 in 1976. In comparison, the national cattle population numbered roughly 2 million in 1974. The Iraqi government supported the country-wide distribution of the Holstein breed by making young Holstein bulls available to cattle owners. Later on, free veterinary services accompanying AI also supported Holstein distribution. Higher-yielding crossbreeds increased demand for AI with Holstein semen (El Dessouky, 1977). In the following decades no more gene flows were reported, as Iraq became completely isolated from the rest of the world. The formation of the Israeli Holstein breed dates back the early 20th century, when imported European cattle were crossbred with locally available cattle (e.g., Damascus cattle). In the 1950s, massive imports of North-American Holstein cattle contributed to a sharp increase in the genetic potential for milk production, which was further enhanced by the strict national breeding plan. This led to the formation of the Israeli Holstein breed, which is known for its high potential for milk production under subtropical and arid conditions and intensive management, and also for the low butterfat and protein content. These developments led to an increased demand for the Israeli Holstein in countries with similar climatic conditions. This is illustrated by the large number of Israeli Holstein cattle exported to Iran, Turkey and Egypt (Israeli Cattle Breeder’s Association, 2001). For more details, see Annex 9.1 A 37, which sums up information from different periods, also reflecting political dependency of export volume and destination. In the tropical and subtropical zone breeding emphasis incorporating exotic breeds is on crossbreeding for milk production but there is also some development of beef. India, Pakistan and Bangladesh are countries in this zone that failed several times to establish pure breeds and made a vast effort to crossbreed taurindicus cattle for milk. Many German and Swiss projects were conducted. Southeast Asia directed its efforts toward breeding dairy cattle, but in many regions - e.g., in the Philippines - they were unsuccessful. Other examples are breeding of Danish Red near Bangkok, German dual-purpose breeds in Thailand, and German Holsteins in Malaysia. In India, there are two prominent examples of successful genetic progress in cattle breeding, Operation Flood and the Indo-Swiss cattle project. Both are characterised by a comprehensive approach, which includes all aspects of the production environment, CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW 265 marketing and training, as well as a long-term commitment by a national or local body. Between 1961 and 1975, bulls for semen production of different dairy breeds (Holstein on the forefront) were imported to India from Australia, Denmark, Canada, New Zealand, United Kingdom and the USA (Katpatal, 1977). In 1985, 1,000 Holstein heifers were imported from Germany and, over many years, Brown Swiss were brought in from Switzerland and the USA. These imports were crossbred with local cattle to form the Sunandini and Karan Swiss breeds under the auspices of the Indo-Swiss cattle project in Kerala State. Although Holsteins were in the forefront of live animals imported into India, the National Dairy Development Board of India (Operation Flood) favoured and used Jersey cattle for crossbreeding because of their desirable high butterfat percentage and small body size of the cows. Since the start of the 19th century, the Philippines have imported breeding animals. The Southwest of the Philippines and parts of Indonesia are somewhat drier and are suitable for cattle ranching. Brahman, American, Brazilian and Australian Zebu crossbreeds and some large-sized European beef and dual-purpose breeds were imported to these areas. The Brahman breed has become the most popular exotic breed due to its adaptability. In Table 5, an overview of importations into the Philippines is given. In the National Medium Term Livestock Development Programme (1993-1998), cattle, sheep, goats, pigs, and other animal species were imported to the Philippines. Purebred Brahman and Simbrah cattle were imported from the USA and kept in nucleus herds. Superior animals were multiplied and genetic improvement reached rural village level through AI or live bull crossbreeding of local cattle (Boadle et al., 1997; Loculan, 2002).

Table 5: Cattle breeding imports into the Philippines

Year Number of Breeds breeds 1900-1935 15 Dairy: Jersey, Shorthorn, Holstein, Guernsey Beef: Galloway, Angus, Devon, Hereford, Sussex Zebu beef: Nellore, Hariana, Bhagnari Zebu dual-purpose: Sahiwal, Red Sindhi, Tharparkar 1946-1970 12 Dairy: Brown Swiss Beef: Brahman, Chiricano, Santa Gertrudis, Ongole, Charolais, Charbray Braford, Red Poll Others: Red Danes, Australian Illawara, Balinese, Madura 1971-1983 9 Beef: Indu-Brazil, Droughtmaster, Beefmaster, Belmont Red, , Chianina, Marchiagiana, Simmental, Maine Anjou 1990 2 Dairy: Australian Friesian Sahiwal Dual-purpose: Simbrah Year Head Breeds 1993 8,816 Simbrah 1994 6,893 Brahman, Simbrah, Holstein, Sahiwal 1995 12,546 Senepol, Nellore, Simmental, Brahman 1996 3,648 Brahman, Senepol

Source: adapted from: Loculan (2002) 266 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW

From 1999 onwards, Indonesia imported bulls of various breeds, including the Simmental, from Australia. The import scheme, intended to increase meat production, was fostered by the Indonesian government. Artificial breeding centres were established in provincial areas, and technicians were trained and equipped with nitrogen, containers, and motorbikes to transport subsidised AI semen to farmers. 20% of Indonesia’s cows were inseminated with Simmental or Limousine semen (Hadi et al., 2002; Meat and Livestock Australia, 2001; Simmental Australia, 2002). Imports and exports of other Asian countries could not be quantified.

3.1.5 Africa

Information on the current status and trends of gene flow in important African countries is limited, despite the significance of cattle breeding in countries such as South Africa, Namibia, Zimbabwe and Zambia in the South, Morocco, Tunisia and Algeria in the North, and Kenya, Tanzania and Uganda in the East. The following section depicts the example of Mozambique, where gene flow has been documented. Holstein cattle were first imported to Egypt in the 1930s. Between 1954 and 1961, imports from the Netherlands and Denmark followed and, from the 1970s, semen was imported from the USA (Nigm, 1990). AI was of practical use in commercial farming in Mozambique as early as 1962. At this time, imported semen was used exclusively. The Africander breed from South Africa was the most significant for beef production. Other breeds used included Hereford, Angus, Sussex, Simmental, Pinzgauer, Charolais, Chianina, Brahman and Santa Gertrudis. In 1974, a first effort was made by governmental bodies of Mozambique to establish an AI centre, which did however not succeed. In 1977, the national AI centre was reopened and a subsidised AI scheme removed the need for fees. In the 1980s, however, it was observed that AI had serious limitations under ranch conditions and its importance diminished. AI was also not introduced in the small-scale sector. Failure of AI resulted from the lack of an established dairy industry and breeding policy. The establishment of AI centres was premature. From a dairy herd of about 3,000 cows in state and private farms in Mozambique in 1983, about 70% were artificially inseminated. Figure 7 illustrates the breeds and countries of origin of the semen used. Almost 95% of the cows were inseminated with Holstein semen. The majority came from Mozambique, and significant percentages came from New Zealand, Denmark and Germany (Rocha et al., 1987). CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW 267

Figure 7: Number of inseminations by breed and origin of semen in the dairy sector in Mozambique 1983

Holstein from Mozambique

492 154 Holstein from New Zealand 212 Holstein from Denmark 104 11 Holstein from Germany

1193 Jersey from Denmark

Brown Swiss from Switzerland

Source: Rocha et al. (1987)

3.1.6 World

FAO surveys conducted by Chupin and Schuh (1993) and Chupin and Thibier (1995) analysed the use of AI and semen trade in developing and developed countries. In 1991, 76% of developing countries included in a survey reported use of AI in cattle in their countries (100% Near East, 87% Asia, 79% Latin America, 57% Africa). In some of these countries, the development of AI programmes was still in experimental stages. 71% of countries using AI also produced semen locally, while the remainder used only imported semen. 65% of the developing countries included in the analysis imported cattle semen. This percentage is highest in Asian countries (78%), lowest in African countries (51%), and intermediate in Latin American (68%) and Near East countries (75%). On average, most imports went to Latin American countries, with almost 121,000 imported semen doses, and the least went to African countries, which imported just 11,750 doses. Asian countries imported 37,437 doses and Near East countries 28,877 doses). Based on data from 62 countries, the number of inseminations in developing nations increased by 131% between 1980 and 1991 (Africa: -5%; Latin America: +11%; Asia: +85%; Near East: 203%). The stagnation in Africa insemination rates is probably due mainly to the unstable political situation in Southern Africa and the declining interest of the western world in Africa following the end of the East-West conflict. In Latin America, where beef is important, AI development was probably constrained by economics. The countries with rapid dairy development in the Near East and Asia have also had rapid AI development. In developed countries the number of inseminations decreased by 13% between 1980 and 1991 (EU: -17%; North America: -5%; Eastern Europe: -29%; Western Europe: -8%; others: +21%). During this period, the use of AI decreased most significantly in dual- purpose breeds (-29%), while AI in dairy breeds decreased by -12% and AI in beef breeds 268 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW increased by +25%. The decline in dairy inseminations in the developed countries is mainly due to yield increases and reduced cow numbers. In general, AI coverage in developing countries increased between 1980 and 1991, through there was considerable variability between countries and regions. In 1991, AI coverage in developing countries was highest in Asian countries, followed by Near East, Latin American and African countries. In all regions AI coverage was highest for temperate breeds and lowest for local breeds, with AI of crossbreds between local and temperate breeds being intermediate. As could be expected, AI coverage in developed countries was higher than in developing countries. Generally, it was highest in dairy breeds and lowest in beef breeds. For more details on AI coverage differentiated for regions and breed groups see Annexes A 38 and A 39. Of inseminations in the surveyed developing countries, 43% were by local breeds, 30% by crossbreds and 27% by temperate zone breeds. These percentages were highly variable by region. In Africa inseminations by the three breed groups had about equal shares, while in Asia inseminations by local breeds clearly dominated. In Latin America, the majority of inseminations were by crossbreds, while in the Near East most were by temperate breeds. In developed countries dairy breeds made up a 61% majority of all inseminations. 23% of inseminations were performed with dual-purpose semen and the remaining 16% with semen from beef breeds. For more details see Annexes A 40 and A 41. The main source of gene flow from developed countries was North America (USA, Canada), which produced 72% of global semen exports (Chupin and Thibier, 1995). The second largest exporter was the EU, from which 19% of semen exports originated. The remaining 10% of exports came from non-EU countries in Eastern and Western Europe and other countries (Australia, New Zealand, and South Africa). Semen exports from the EU in 2003 point in a similar direction except that numbers were lower (as the EU is the only trading partner for the importing countries considered) and that mean numbers were highest for countries of West Asia/Maghreb followed by Latin America (which mirrors geographical factors). For data on semen exports from the EU in 2003 differentiated by country of destination, refer to Annex 9.1 A 26. 88% of semen imported by developing countries globally was from dairy breeds, which made up 98% of imports in the Near East, 73% in Asia, 87% in Africa, and 91% in Latin America. Among the dairy breeds, the Holstein breed was clearly dominant, accounting for 86% of all imported semen doses, with much smaller shares contributed by the Jersey breed (4%) and other dairy breeds (mainly Brown Swiss). 9.5% of all imported semen doses were from beef breeds and 2.1% from tropical breeds (mainly Brahman, Red Sindhi and Sahiwal) (Annex 9.1 A 30). Hemme (1995) presents a study of the dynamics of semen trade in relation to trade values from 1983 to 1993. He found that the value of semen traded internationally increased almost steadily and, in total, tripled over this period. Since then, almost all countries have increased their exports in terms of absolute numbers. Proportionally, the market share of US exports shrunk to the benefit of EU exports. Within the EU, Germany, the Netherlands and France dominated exports. Mexico, Brazil and Venezuela were the main importers among southern countries and all showed constant growth rates, according to Hemme. In Asia, Japan and South Korea extended their imports, while in other parts of the South no major changes occurred. The highest prices were paid for semen in Japan and in the EU, CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW 269 indicating the interest of customers in the top dairy cattle genetics of the temperate zone. Relatively low prices were paid in Latin America, reflecting the main use of semen in commercial herds. Philipsson (2002) analysed four cases of cattle breeding systems with AI in less developed countries (Pakistan, Kenya, Sudan, West Indies). The author identified problems in the following areas: • Breeding objectives and utilisation of genetic resources; • Methodological, technical and operational issues; • Policy and organisational aspects. AI is a useful technology for the breeding of dairy cattle, but is often of marginal economic benefit in beef cattle breeding. The problems with dairy cattle in the humid tropics is the lack of selected breeds or problems with excessive upgrading with temperate zone breeds.

3.2 Gene flow indicated by changes in the breed composition In this section, examples of changing breed composition in national cattle populations are given, focusing primarily on the proportion of the Simmental, Holstein and Brahman breeds. In some traditional Simmental-keeping regions of central Europe, the significance of the breed waned in the second half of the 20th century. The Simmental population decreased and was partly crossed with the Red Holstein. Table 6 gives an illustration from the 1960s, when AI began to gain importance, to the beginning of the 1980s in Switzerland. Because of the very high milk prices relative to beef, dairy farmers in the plains started to upgrade the Simmental to the Red or Black Holstein, farmers at medium altitude continued to breed Simmentals from a Red Holstein x Simmental foundation, and only farmers at higher altitude maintained the Simmental without gene flow from the Holstein (Piot, 1987).

Table 6: Total numbers of cattle by breed in Switzerland

Total numbers of cattle by breed in Switzerland in 1,000 1966 19731978 1983 Change 1966 to 1983 in % Brown cattle 852 894 884 786 -8 Red Pied 885 894 897 851 -4 (Simmental) Black Pied 30 86 190 229 660 (Holstein) Eringer 20 15 15 14 -31 (Herens) Crosses 10 22 37 52 439 total 1,796 1,911 2,024 1,933 8 Switzerland

Source: Piot (1987) 270 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW

In the 1970s, the former Soviet Union decided to pursue Holsteinization. While the Simmental had been the leading cattle breed in the former USSR, no further imports of Simmentals were made and large-scale upgrading was practised following the decision to convert to Holstein. The number of Simmental cattle remained almost constant between 1974 and 1980, but their share of the national cattle population decreased from 29.2 to 25.6% due to the increased numbers of Holsteins (Bagrii, 1982). Although the Simmental was rehabilitated by Russia as an accepted breed in 1997, it had lost its importance. According to information provided by the Russian Cattle Breeders Federation Rosplemobje-dinenje for 2001, the share of Simmental pedigree cattle had declined to about 10%. Even within nominally Simmental herds, the use of Red Holstein bulls caused the cattle to look more like Red Holsteins than Simmentals. In Botswana in the 1980s, extension services made recommendations on the use of exotic cattle breeds. They stated that, due to the large genetic differences between the local Tswana cattle and exotic (Bos taurus) breeds, higher heterosis effects could be expected when crossbreeding Tswana with exotic rather than local breeds. Exotic live animals originated mainly in South Africa (Country Report of Botswana, 2003). The Tswana breed is a Sanga cattle type introduced in the late 18th century to Botswana from Eastern Africa (Felius, 1995; Payne and Hodges, 1997). In the 1980s, government institutions started to promote the use of exotic bulls (including Simmental) on Tswana cows by subsidies and the operation of an AI scheme. This led to a decrease in the share of purebred Tswana cattle in the national cattle population from 75% in 1983 (Felius, 1995) to 50% in 1996, and to decreased breeding efforts in Tswana cattle (Nsoso and Morake, 1999). The diagram in Annex 9.1 A 42 illustrates that between 1987 and 1995 almost 95% of the bulls used in Botswana were exotic breeds (Brahman: 57%; Simmental: 16%; Charolais: 13%; other exotics: 9%; local breeds: 5%). However, under small-scale livestock-keeping conditions pure Tswana cattle are reported to economically outperform the Simmental- Tswana-crosses (Country Report of Botswana, 2003). Additionally, the use of exotic bulls under small-scale livestock-keeping conditions is questionable due to lack of mating control, management expertise, supplementary feeding and veterinary inputs (Nsoso and Morake, 1999). Milk production is not a major production goal in Botswana. According to the Australian Brahman Breeders’ Association (2004), minimal numbers of Brahman cattle were first introduced to Australia in about 1900, though there are also reports of earlier imports (Payne and Wilson, 1999; Vercoe, 1989). The 1930s, the 1950s, and - most notably - the 1970s, brought subsequent major cattle imports from the USA. The breed’s resistance to different stress factors (e.g., heat, ticks) led to its spread in tropical and subtropical regions of Northern Australia. Figure 8 illustrates the effects of the introduction of the Brahman, steady increase in total number as well as in the share of zebu cattle breeds in Queensland/Australia. While the share of zebu cattle breeds in Queensland/Australia was zero until 1930, it grew to 4% in 1950, reached more than 50% at the end of the 1970s and was expected to be close to 100% in the late 1990s. Breeding in marginal areas which had not previously been used for beef production was allowed by the introduction of tropical cattle breeds such as the Brahman and its derivates. For more recent development of cattle registrations by breed in Australia, refer to Annex 9.1 A 43 (Croaker, 2002; Daly, 1981; Griffith et al., 2003; Payne and Hodges, 1997; Vercoe, 1989). CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW 271

Figure 8: Number of cattle in Queensland/Australia

12 Bos taurus Bos indicus and crosses 10

8 5.9 4.0 6 0.7 0.2

4 5.6 4.7 5.2 5.0 2 4.1 4.4 number of animals in 1'000'000

0 1900 1930 1950 1965 1973 1977

Source: own elaboration, data from Daly (1981)

In a development similar to that in Australia, the share of Brahman cattle in South Africa grew from 4.4% to almost 57% between 1965 and 1985. The increase was mainly at the expense of the Africander breed, an indigenous South African Sanga-type breed kept as the main beef breed by white farmers in drier areas (Brahman Cattle breeders’ society South Africa, 2004). The number of breeders by breed in 2003 is given in Annex 9.1 A 44. Brahmans were imported primarily to South Africa and Namibia, and moved from there to Botswana. The use of the Simmental breed varies from country to country dependent on the primary purpose of the breed in the country. This is due to dependencies on the level of milk prices (Switzerland) and on the milk-beef-price ratio. In traditional Simmental-breeding countries in Central and Eastern Europe, the breed is used as a genuine dual-purpose breed for milk and beef production. In contrast, in the New World and in North-West Europe where the breed has only been recently introduced, it is mainly used for beef. Because of more attractive prices for milk than for beef, the Swiss Simmental and the French Montbéliard sub-type of the breed were oriented more towards milk (see also Annex 9.1 A 45).

3.3 Gene flow indicated by the introduction of foreign genetic material into existing breeds The gradual introduction of foreign genes into existing breeds is another indicator of gene flow. Although its impact and danger are higher, it is much more difficult to detect and measure than the change of national composition of animal populations. The introduction of foreign genes into existing breeds is illustrated by the example of the Simmental breed in selected European countries. Ayrshire bulls were introduced into Simmental breeding in Southern Germany in the 1950s to improve milk yield and udder quality, but this did not lead to sustained results. In 1984, a change in the breeding policy following the introduction of the EU milk-quota system put an end to widespread crossbreeding schemes of Simmental with Red Holstein in Germany, although some Red Holstein or half Red Holstein bulls were still used (Aumann, 2003; Averdunk et al., 2001). 272 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW

Ayrshires exerted a strong influence on the Simmental breed the Czech Republic. The introduction of imported Red Holstein blood from North America in the 1970s led to the expected higher milk yields in Simmentals, but to the detriment of parameters important for beef production. The milk-beef-price ratio drove this development. In contrast to that, the influence of Red Holstein introduction on the Simmental breed has been more pronounced in Switzerland (first crosses in 1967). In Russia, the introduction of Ayrshire and Montbéliard genetic material into the Simmental breed led to increased milk yields per cow in the 1970s (Bagrii, 1982). From the 1980s, increased introduction of Red Holstein germplasm is reported (Sonn, 1985). In the early 1990s, these developments were questioned and a return to typical dual-purpose and pure Simmentals was requested (Sereda et al., 1991). However, the Simmental share of the national herd had been reduced from 40% to 10% by 2001 because of the Holsteinization policy. Introduction of Red Holstein genetic material has also been a general policy in other East European Simmental (Felius, 1995). In the 1970s, the Montbéliard breed in France, a sub-type of the Simmental, had no noteworthy influence of foreign genetic material. Starting in the early 1980s, however, Holstein germplasm from the USA was introduced and increased the share of foreign genetic material to 4%. For genetic material from Canada this value was 1% in the early 1990s. In the French Simmental breed the share of Swiss Simmental genetic material grew from 15% in 1970 to about 30% in 1992, reaching a peak of almost 45% in the mid 1980s. As the Swiss Simmental was influenced by Canadian and American Red Holstein genetic material, the share of Holstein material in the French Simmental increased too. The share of German genetic material, which contained much less Holstein influence, was zero in most of the 1970s. The share of foreign genetic material was much larger in the French Holstein breed, which was almost completely absorbed by foreign Holstein germplasm through embryo imports. The percentage of US American genetic material in the breed grew from about 5% in the 1970s to almost 80% in 1992. The share of Canadian genetic material in the same period was between 10% and 25%, with a peak in 1980. The share of Dutch material decreased from much higher levels in 1970 to less than 10% in 1992 (Boichard et al., 1996). In 1995 the estimated percentage of Holstein genetic material in Black Pied cattle populations averaged 66% in the EU, 59% in other European countries, and 36% in the former USSR countries (Felius, 1995). AI, which made these developments possible, caused the active breeding population of Holsteins, if considered genetically, to be rather small (Wickham and Banos, 2004). As the dairy cattle breeding industry has become more globalised and genetic material of temperate breeds - mainly Holstein - has been increasingly traded internationally, lack of consideration of genotype-environment interaction has caused sires of temperate origin to be genetically overestimated when used in tropical and subtropical countries (Holmann et al., 1991; Stanton et al., 1991). In 1984 the International Bull Evaluation Service (Interbull) was founded by the International Dairy Federation, the European Association for Animal Production and ICAR (at the initiative of Professor Gravert), at the onset mainly for better comparability and transparency within the temperate zone. Although virtually all data come from intensive dairy systems that do not provide breeding value estimates in most developing countries, it CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW 273 has helped exporters to purposefully target potential markets and importers to have a better information basis for their buying decisions (Banos and Sigurdsson, 1996; Smith and Banos, 1991; Zwald et al., 2003). However, the international exchange of genetic material from few sires has also resulted in increasing levels of inbreeding since the early 1980s, prompting concerns about and loss of genetic variation (Young and Seykora, 1996). In some cases, introduction of foreign genetic material takes the form of continuous gene flow into local breeds, leading to their absorption and replacement by new breeds. As an example, the gradual introduction of the Simmental breed through upgrading or expansion of imported purebred herds has led to the replacement of local breeds or strains in some parts of the world, most notably in Europe. For data on the Simmental breed, see Table 7, and for other breeds, see Annexes A 46 and A 47. In many countries the Simmental itself is now facing increasing absorption by other breeds, especially by the Holstein, which performs better in different environments but is also being introduced to production conditions where it cannot express its performance potential (see Annex 9.1 A 48). 274 CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW

Table 7: Examples of replacement of local cattle by imported Simmentals

Absorbed Absorbing Place Time Aspects Source breed breed Local cattle Simmental Europe Late 19th Share of Simmental Felius, strains and early increased and local cattle 1995 20th strains absorbed century Vogtländer Simmental Germany Early Governmental regulations: Muller, 20th traditional Vogtländer 1993 century breeding area to be changed Hinterwälder Simmental Germany Second Superior efficiency Furthmann, half of 1987 20th century Limpurger Simmental Germany 20th Superior efficiency Haller, century 2000 Chianina and Simmental Italy 1960s and Replacement of local Averdunk Romagnola 1970s draught breeds et al., 2001 among (mechanisation of others agriculture) Hereford, Simmental USA 1970s Spread of Simmental Peterson, Angus, diminished relative 1988 Shorthorn, importance of traditional Charolais beef breeds East Simmental, Turkey 1980s Crossings implied a Oklahoma Anatolian Brown Swiss reduction of purebred East State Red Anatolian Red cattle. University, 2004 Tswana Simmental Botswana 1980s and Reduction of purebred local Nsoso and cattle among other 1990s Tswana cattle Morake, exotic breeds 1999 Istrian cattle Brown Croatia 1990s AI with exotic breeds, Dasovic, Swiss, superior efficiency 1992 Simmental and others

3.4 Gene flow indicated by Country Reports The excerpts of the Country Reports provided the most extensive information on cattle. However, the summary provided in Table 8 does not give a representative picture of the situation in different regions. The largest inflows are reported in the whole of Europe, followed by Western Asia. Large imports may also be expected from Latin America, but the analysis was based on only one country report excerpt, which cannot be considered representative. The same holds true for the outflows mentioned in the excerpts. The flows out of Western and Northern Europe, for example, are reported to be medium. However, countries known to have large outflows, such as France, Germany, the Netherlands, Italy (semen), Austria, Denmark or the UK, are not included. CURRENT STATUS AND ACTUAL TRENDS OF CATTLE GENE FLOW 275

The breeds mentioned in the excerpts should also be interpreted cautiously. Discussion of the same breed in different regions may have very different implications. This is the case regarding the Brahman and Holstein breeds in Africa and Latin America: the Brahman is a beef breed useful for beef-ranching in dry tropical or outlying areas while the Holstein is a dairy breed used for high potential, high feeding-level environments. Additionally, it is likely that important information is missing from the analysis of excerpts. It is well documented that there was an important inflow of Holstein genes into Oceania, however this is not mentioned in the present analysis because only excerpts from Cook Island, Palau, Samoa and Vanuatu were available and only the excerpt from Samoa contained information. The excerpts for cattle are summarised in Annex 9.1 A 49.

Table 8: Country Reports on gene flow of cattle

Breed Extent of gene flow Main breeds Breed Disadvantages Country reports in and out of the imported replace- of exotic breeds containing information region ment In Out Africa Medium low Brahman, Holstein, medium - Not named Botswana, Burkina Faso, - high Holstein-Friesian, high Burundi, Cameroon, Jersey Eritrea, Ethiopia, Ghana, Malawi, Zambia, Zimbabwe Eastern and high - medium Holstein, medium - Replacement of Albania, Belarus, Bosnia- Southern very Simmental high local breeds, Herzegovina, Bulgaria, Europe high failure to Czech Republic, Russian upgrade Federation, Serbia and production Montenegro, Slovakia, systems Slovenia Latin Medium 0 Brahman, Holstein medium Not named El Salvador America North High high Holstein low Not named Canada, United States of America America Oceania High 0 Braford, Brahman, 0 Not named Samoa Droughtmaster Southern High 0 - low Holstein, Holstein- low Lack of Bhutan, China, and Eastern Friesian, Jersey adaptation to Indonesia, Japan, Laos, Asia local production Myanmar, Nepal, conditions Pakistan, Philippines, Sri Lanka, Tajikistan, Uzbekistan, Vietnam Western and high - medium Holstein high Not named Finland, Ireland, Sweden, Northern very United Kingdom Europe high Western High low Holstein medium Loss of local Armenia, Azerbaijan, Asia breeds Turkey 276 REFERENCES GLOBAL GENE FLOW OF CATTLE

4 REFERENCES GLOBAL GENE FLOW OF CATTLE

Agro-News Aktuelles aus der Landwirtschaft. 2003. Erfolgreiche Schweizer Viehexporteure. Landwirtschaftlicher Informationsdienst (LID). http://www.landwirtschaft.ch/deutsch/aktuell/newsdetail.cfm?ObjID=2AEAC2F0- BAE1-4484-8A093101A63603D7&language=de. (09.07.2004) Alba, J.D. 1978. Progress in the selection of the Latin American Dairy Criollo. World Animal Review 28:26-30. Allen, J. 2002. Country Report: Australia. ACIAR Proceedings Series Number 108: Development strategies for genetic evaluation for beef production in developing countries. 13-22. American Simmental Association. 2004. History of the Simmental Breed. http://www.simmental.org/General%20Information/SIM%20History.htm. (01.07.2004) Arbeitsgemeinschaft Deutscher Rinderzüchter e.V. 2002. Rinderproduktion, Zucht, Besamung, Leistungsprüfung in der Bundesrepublik Deutschland. Jahresberichte 1972-2002. Arbeitsgemeinschaft Österreichischer Fleckviehzüchter. 2004. Personal communication. Arbeitsgemeinschaft Süddeutscher Rinderzucht und Besamungsstationen e.V. 2002. Jahresberichte. 1998-2002. Asociación Colombiana de Criadores de Ganado Simmental y sus cruces-AsoSimmental. 2004. Simmental. http://www.unaga.org.co/asociados/simmental.htm. (11.07.2005) Aumann, J. 2003. Fleckvieh in Europa: Aktuelle Situation und Zukunftsaussichten - Zusammenfassung des Vortrags auf der Tagung der Europäischen Vereinigung für Fleckviehzucht. Belgrad, September 2003. Zuchtwahl und Besamung 150:28-30. Australian Brahman Breeders’ Association. 2004. A Brief History. http://www.brahman.com.au/aboutbrahmans/history.htm (11.07.2005). Averdunk, G., A. Gottschalk, M. Putz, and E. Rosenberger. 2001. Fleckvieh - Entwicklung zu einer Weltrasse. Arbeitsgemeinschaft Süddeutscher Rinderzucht- und Besamungsorganisationen e.V. München. BLV Verlagsgesellschaft. Bagrii, B. 1982. Problems of utilising genetic resources in cattle breeding. Molochnoe i Myasnoe Skotovodstvo 12:30-35. Banos, G. and A. Sigurdsson. 1996. Application of Contemporary Methods for the Use of International Data in National Genetic Evaluations. Journal Dairy Sciences 79:1117- 1125. Boadle, R., M. Coombe, and J. Croaker. 1997. Pedigree indexing of imported American Brahman breeder cattle in the Philippines. Asian-Australasian Journal of Animal Sciences 10:614-620. Boichard, D., L. Maignel, and E. Verrier. 1996. Analyse généalogique des races bovines laitières francaises. Productions Animales 9:323-335. REFERENCES GLOBAL GENE FLOW OF CATTLE 277

Brahman Cattle Breeders’ Society South Africa. 2004. Background. http://www.brahman.co.za/main.htm (11.07.2005) Chupin, D. and H. Schuh. 1993. Survey of the present status of the use of artificial insemination in developing countries. World Animal Review 74/75:26-35. Chupin, D. and M. Thibier. 1995. Survey of the present status of the use of artificial insemination in developed countries. World Animal Review 82:58-68. Cossee, O. 2003. Emergency Farm Reconstruction Project Kosovo-Analysis of the Final Survey. FAO, Rome, Italy. Country Report of Botswana for FAO’s State of the World’s Animal Genetic Resources. 2003. Country Report of Brazil for FAO’s State of the World’s Animal Genetic Resources. 2003. Ministerio da Agricultura, Pecuaria e Abastecimento. Cowert, C. 1983. Brahmans: at home on any range. Foreign agriculture / Foreign Agricultural Service 21:11-12. Croaker, J. 2002. The influence of Brahman Genetics in the Australian Cattle Industry. ACIAR Proceedings Series Number 108: Development strategies for genetic evaluation for beef production in developing countries. 154-161. Daly, J.J. 1981. Beef cattle breeds. Queensland Agricultural Journal 107:28-50. Dasovic, N. 1992. Stanje Govedarstva I Umjetno Osjemenjivanje Na Prodrucju Istre. Veterinarska Stanica 23:15-19. Dodenhoff, J. and C. Egger-Danner. 2004. 10 Jahre Eurotest Verknüpfungen. Fleckvieh Zucht Besamung Management Markt 11: 46-47. El Dessouky, F.I. 1977. The development of artificial insemination in Iraq. World Animal Review 27:14-18. Epstein, H. and I.L. Mason. 1984. Cattle. In: Mason, I.L. (ed.) Evolution of domesticated animals. Longman, London and New York. Eurostat. 2004. EUROSTAT-COMEX Databank. European Union, Luxemburg, 2004. FAO. 2004. Global Livestock Production and Health Atlas (GLiPHA). 1[1]. 2004. FAO, Rome, Italy. Felius, M. 1995. Cattle breeds: an encyclopedia. Doetinchem: Misset. Fraga, A. 2004. Personal communication. Furthmann, G. 1987. Die Hinterwälder des Schwarzwaldes - Eine erhaltenswerte Rasse. Tierzüchter 39:249-251. Griffith, G.R., R.J. Farquharson, S.A. Barwick, R.G. Banks, and W.E. Holmes. 2003. Estimating returns from past investments into beef cattle genetics RD&E in Australia. 25th International Conference of Agricultural Economists (IAAE), 16-22 August 2003, Durban, South Africa. 1338-1343. Grupp, T. 2003. Spreading the best Fleckvieh genetics to the world. Fleckvieh World 14- 17. 278 REFERENCES GLOBAL GENE FLOW OF CATTLE

Hadi, P.U., N. Ilham, A. Thahar, B. Winarso, D. Vincent, and D. Quirke. 2002. Improving Indonesia’s beef industry. ACIAR Monograph No. 95, vi + 128p. Haller, M. 2000. Seltene Haus- und Nutztierrassen. Stocker Verlag, Graz. Haluska, J. 2002. Osvrt Na Stanje Hrvatskog Agrara (Program Razvoja Obiteljskih Gospodarstava). Stocarstvo 56:43-58. Hanotte, O., D.G. Bradley, J.W. Ochieng, Y. Verjee, E.W. Hill, J.E.O. Rege. 2002. African pastoralism: Genetic imprints of origins and migrations. Science 296:336-339. Hemme, T. 1995. Marktführer USA bekommt zunehmend Konkurrenz. Typex Magazine 63: 63-64. Holbert, T. 1983. Brangus: the making of a genetic blend. Foreign agriculture / Foreign Agricultural Service 21:9. Holmann, F.J., R.W. Blake, R.A. Milligan, P.A. Oltenacu, R. Barker, and M.V. Hahn. 1991. Comparative economic returns from artificial insemination and natural service in straightbred and crossbred Holstein herds in Venezuela. Journal of Dairy Science 74:665-677. Israeli Cattle Breeder’s Association. 2001. Milestones in the History of Israeli Dairy Cattle. http://www.icba-israel.com/icba-milestones.html (11.07.2005). Israeli Cattle Breeder’s Association. 2004. Breeding. http://www.icba-israel.com/icba- breeding.htm (11.07.2005) Jakopovic, I., Z. Dominikovic, M. Bolic, and J. Culi. 1995. Doprinos Roditelja Novih Generacija Bikova Genetskoj Izgradnji Goveda Simentalske Pasmine. Stocarstvo 49:261-268. Katpatal, B.G. 1977. Dairy cattle crossbreeding in India. World Animal Review 22:15-21. Künzi, N. and G. Stranzinger. 1993. Allgemeine Tierzucht. 1. Ulmer Verlag, Stuttgart, Germany. Liang, S.I. 1988. Fleckvieh in aller Welt. China. Tierzüchter 40:384. Loculan, D. 2002. Genetic evaluation of beef production in the Philippines. Development strategies for genetic evaluation for beef production in developing countries 72-80. Machugh, D.E. 1996. Molecular Biogeography and Genetic Structure of Domesticated Cattle. Dissertation. University of Dublin. Machugh, D.E., D.G. Bradley, and C. Meghen. 1994. Genetic characterization and West African cattle. World Animal Review 78. Martínez, R.D., E.N. Fernández, R. Género, and F.J.L. Rumiano. 2000. El ganado Bovino Criollo en Argentina. Archivos de Zootecnia 49:353-361. McDonald, J.C. 1973. Honduras buys U.S. beef breeding stock mostly Brahman. Foreign agriculture / Foreign Agricultural Service 11:11. McDowell, R.E. 1992. Dairy production in developing countries. Proceedings of the New Zealand Society of Animal Production 52:3-5. Meat and Livestock Australia. 2001. Aussie bulls for Indonesia. http://www.mla.com.au/ content.cfm?sid=25&newsid=499&archive=show (11.07.2005). REFERENCES GLOBAL GENE FLOW OF CATTLE 279

Medic, D., A. Antov, and G. Antov. 2002. Genetic Improvement Program for Holstein- Frisian cattle breed in Vojvodina. 7th Symposium on Animal Reproduction, 4-7 September 2002, Ohrid, Macedonia. Meyn, K. 2005. Personal communication. Monke, D. 1999. Export of Semen and Embryos in a Changing Environment. National Conference on Animal Health and International Trade. 83rd LCI Annual Meeting, 29 March-1 April 1999, Nashville, Tennessee. Muller, B. 1993. Vogtland cattle. Neue Landwirtschaft No. 4, 66:66. Neser, F.W.C., J.B. van Wyk, and G.J. Erasmus. 2002. Simmentaler in Southern Africa - A Scientific Review. http://www.simmentaler.org/Publikasies/SIMMENTALER %20IN%20SOUTHERN%20AFRICA.htm (11.07.2005). Nigm, A.A. 1990. Dairy performance of some exotic breeds at Tahreer province. Veterinary Medical Journal Giza 38:351-361. Nsoso, S.J. and T.G. Morake. 1999. A critical look at the use of exotic bulls in traditional beef farming in Botswana. South African Journal of Animal Science 29:100-104. Oklahoma State University. 2004. Breeds of Livestock - East Anatolian Red. http://www.ansi.okstate.edu/breeds/cattle/eastanatolianred/index.htm (11.07.2005). Orbita. 2004. El Simmental en Colombia. http://www.homestead.com/bonavet/files/ page14.html (11.07.2005). Payne, W.J.A. and R.T. Wilson. 1999. An introduction to animal husbandry in the tropics. 5th Edition. Blackwell Science, Oxford. Payne, W.J.A. and J. Hodges. 1997. Tropical cattle: origins, breeds, and breeding policies. 1st Edition. Blackwell Science, Oxford. Peterson, E.B. 1988. Fleckvieh in aller Welt. Nordamerika. Tierzüchter 40: 380-381. Philipsson, J. 2002. Sustainability of dairy cattle breeding systems utilising artificial insemination in less developed countries - examples of problems and prospects. In: ILRI-SLU 2003. Animal Genetics Training Resource CD-ROM. Version 1. Mwai O. and B. Malmfors (eds.). ILRI (International Livestock Research Institute), Nairobi, Kenya and SLU (Swedish University of Agricultural Sciences), Uppsala, Sweden. Piot, J.C. 1987. Die Zukunft unserer Rinderrassen in der Schweiz. Simmentaler Fleckvieh 3:2-9. Qiu Huai, Ju Zhiyong, and Chang Zhijie. 1993. A survey of cattle production in China. World Review Animal 76. Rocha, A., C. Concha, and D. McKinnon. 1987. Mozambique. Artificial insemination of cattle. Some aspects of breeding policy. World Animal Review 64:12-20. Rusch, R.E.F. 1988. Fleckvieh in aller Welt. Südliches Afrika. Tierzüchter 40:383-384. Schahidi, R., H. Mahmmudzadeh, R. Kasravi, M. Klopcic, K. Malinger, N. Siard, and S. Zgur. 2001. Breeding, artificial insemination and recording of cattle in Iran. 32nd ICAR and INTERBULL Meeting, 16-20 May 2000, Bled, Slovenia. 10:113-117. 280 REFERENCES GLOBAL GENE FLOW OF CATTLE

Schuhmacher, R. 1983. Beefmaster: the raiser’s edge. Foreign agriculture / Foreign Agricultural Service 21:10-11. Schweizerischer Fleckviehzuchtverband. 2000. Geschäftsbericht 1999/2000. Schweizer Fleckvieh 7:5-27. Sereda, L., J. Luttmann, W. Nechin, and H. Sonn. 1991. Fleckvieh im Ausland. Tierzüchter 43:284-285. Simmental Australia. 2002. Simmentals Double Beef Production in Indonesia. http://www.simmental.com.au/news_items/default_news_50.html (11.07.2005). Smith, C. and G. Banos. 1991. Selection within and across populations in livestock improvement. Journal of Animal Science 69:2387-2394. Sonn, H. 1984. Aspekte der tierischen Erzeugung in Afrika - Erfahrungen beim Einsatz Europäischer Rinderrasen in Afrika. Tropenlandwirt 19:96-105. Sonn, H. 1985. Deutsches Fleckvieh im Ausland-Ein wichtiger Beitrag zur Entwicklung der Welt-Fleckviehzucht. Tierzüchter 37:101-102. SOPEXA. 2004. The French Simmental Breed. http://www.elevage-francais.com/ang/ Bd.asp?Id=18 (11.07.2005). Stange, J.D. 1988. Fleckvieh in aller Welt. Südamerikanischer Tierzüchter 40:381-382. Stanton, T.L., R.W. Blake, R.L. Quaas, and L.D. Van Vleck. 1991. Response to selection of US Holstein sires in Latin America. Journal of Dairy Science 74:651-664. Vaccaro. 1990. Survival of European dairy breeds and their crosses with zebu in the tropics. Animal Breeding Abstracts 58:475-494. Vercoe, J. 1989. Animal adaptation The development of the beef industry in Northern Australia. . Special Feature - International Agriculture: 50-56. Wickham, B.W. and G. Banos. 1998 Impact of International Evaluations on Dairy Cattle Breeding Programmes. 6th World Congress on Genetics Applied to Livestock Production, 12-16 January 1998, Armidale, Autralia. 23:315. World Bank. 2003. Implementation completion report on a trust fund grant in the amount of $US18.4 Million to the United Nations interim administration mission in Kosovo for the benefit of Kosovo, Serbia and Montenegro (Kosovo) for an emergency farm recontruction project. 26932. http://www-wds.worldbank.org/servlet/ WDSContentServer/WDSP/IB/2004/01/08/000090341_20040108095652/Rendered/ PDF/269320YF.pdf (11.07.2005). Young, C.W. and A.J. Seykora. 1996. Estimates of Inbreeding and Relationship Among Registered Holstein Females in the United States. Journal Dairy Sciences 79:502- 505. Zwald, N.R., K.A. Weigel, W.F. Fikse, and R. Rekaya. 2003. Identification of factors that cause genotype by environment interaction between herds of Holstein cattle in seventeen countries. Journal of Dairy Science 86:1009-1018.

ANNEX 281

ANNEX

9.5 Global gene flow of pigs

K. Musavaya, M. Mergenthaler, A. Valle Zárate 282 TABLE OF CONTENTS

TABLE OF CONTENTS

List of tables 283

List of figures 283

Abbreviations 283

1 Introduction 284

2 Historical development of pig gene flow 284 2.1 Gene flow during pig domestication and breed formation 284 2.2 Influence of breeding methods in further development and spread of pig breeds 285

3 Development of pig gene flow after World War II 286 3.1 Influence of technology and global mobility on the spread of breeding pigs and changes in breeding organisation 286 3.2 Main distributors of breeding pigs 287 3.3 Main recipients of breeding pigs 290

4 Current status and actual trends of pig gene flow 292 4.1 Gene flow indicated by selected export and import data for live breeding pigs 292 4.1.1 European Union 292 4.1.2 North America 298 4.1.3 Asia 298 4.2 Gene flow of pigs indicated by the introduction of foreign genetic material in existing breeds 299 4.3 Pig Gene flow indicated by Country Report excerpts 301

5 References global gene flow of pigs 303

LIST OF TABLES 283

LIST OF TABLES

Table 1: Early PIC joint ventures, daughter companies and associates worldwide 291 Table 2: Mean annual EU imports of purebred breeding pigs 1992-2003 by origin 297 Table 3: Pig breeding material imports to Thailand 299 Table 4: Country Report excerpts on gene flow of pigs 302

LIST OF FIGURES

Figure 1: EU exports of purebred breeding pigs (highlighting the Netherlands) by destination and share of extra-EU exports 293 Figure 2: Annual EU purebred breeding pig export by destination 1992-2003 295 Figure 3: EU imports of purebred breeding pigs by destination and share of extra-EU imports 1992-2003 296

ABBREVIATIONS

AI Artificial insemination BC before Christ EU European Union EU-15 Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Portugal, Spain, Sweden, The Netherlands, United Kingdom FAO Food and Agriculture Organisation of the United Nations GDR former German Democratic Republic JSR John S. Rymer breeding company NPD Northern Pig Development Company, renamed National Pig Development Company (UK) PIC Pig Improvement Company Limited (UK) UK United Kingdom USA United States of America

284 INTRODUCTION

1 INTRODUCTION

The trade with breeding pigs cannot be quantified easily because commercial companies dominate the exchange market of some major countries and details are normally confidential, unless a company uses figures in advertisements. Therefore this chapter is based on interviews with Dr. Maurice Bichard, a senior expert and a retired leading scientist of the former Pig Improvement Company Limited (PIC), and Mr. David Steane who provided additional information for Southeast Asia. Isolated statistics of Southeast Asia and Canada and the international Eurostat database have been used for trade of the EU-15 countries. However, the latter database lacks information on breeds. Additionally, some emphasis has been put on Large White and Duroc breeds to trace typical global gene flows. The Large White includes very diverse types induced by its particularly high adaptability to wide-ranging conditions. It adapted in temperate regions as well as in the tropics. It is usually used in crossbreeding programmes as a maternal line. The Duroc is a meat-type pig with excellent adaptability and good combination characteristics. These features have led to the wide spread of Duroc to the tropics, its use in crossbreeding programmes in both paternal and maternal lines, and as an exotic replacing local populations.

2 HISTORICAL DEVELOPMENT OF PIG GENE FLOW

2.1 Gene flow during pig domestication and breed formation The world pig population today accounts for 948 million pigs. 284 million, about 30 % are found in developed countries, the remaining 70 % in developing countries (FAO, 2004). The ancestors of the domesticated pig are found among the wild species Sus scrofa and Sus vittatus (Comberg, 1978). Although there are conflicting theories about where pigs were first domesticated, the earliest known remains of domesticated pigs have been found in Southeast Anatolia and were dated 7,000 BC (Epstein and Bichard, 1984). Several domestication centres have been identified. The earliest one is assumed to be in East Asia (China) and a later one in the area of the Mediterranean and Baltic Sea (Kräußlich and Brem, 1997). From the domestication centre in Asia domestic pigs spread to Europe and to North and Northeast Africa. In Europe the domesticated pig was first introduced into the Caucasus and Balkans from south-west Asia in the second half of the 7th millennium BC, which is slightly later than in Anatolia (Epstein and Bichard, 1984). Recent work by Larson et al (2005) reveals a different pattern that involves six possible domestication centres across Eurasia. They suggest that modern European pigs stem from European, rather than Near-Eastern, wild boars. In the 15th century the Portuguese brought pigs from China and Southeast Asia to Europe. In the 16th century Portuguese and Spaniards introduced pigs into Central and Latin America. Local domestic pigs in Britain were crossed with imported Chinese, Southeast Asian, Portuguese and Neapolitan pigs, to create the modern European breeds. They - and pigs from other sources - were imported into North America. HISTORICAL DEVELOPMENT OF PIG GENE FLOW 285

In Europe breed formation before the 18th century is not clear. Pigs were long unfashionable and breeding records were not considered important. In Britain pigs were imported from China and Indo-China directly or via Naples and Portugal in 1770-80. More productive breeds, not distinctly different, were developed, and small herds were established. Inbreeding subsequently created distinct types and, when successful, the need for more animals and a herdbook arose. The first herdbooks for pigs in Britain were created in 1884 (Epstein and Bichard, 1984). The largest national pig populations today can be found in China, USA, Brazil, Germany and Spain in descending order. Highest pig concentrations (head/square km) can be found in Tuvalu, Singapore, Netherlands, Denmark and Malta (FAO, 2004).

2.2 Influence of breeding methods in further development and spread of pig breeds After local breeds emerged all over Western Europe, groups of breeders formed herdbooks and breeding associations in the late 19th and early 20th centuries. These followed the principles of the early livestock improvers and promoted the virtues of their particular breeds. Purebreeding was normal and individual farmers sourced their boar from the most successful farmers in the region. In the 1930s Mendelian genetics was combined with statistical methods to produce a theoretical basis for animal improvement. This methodology has gradually taken over from traditional methods as the driving force for pig improvement. Additionally moved away from small-scale farms with one or two sows towards large specialised ventures, later equipped with modern technology and commercially orientated. Although the pace of this change varies from region to region, a general trend to more concentrated pig production can be seen worldwide (Glodek and Bichard, 1994). Since the late 1970s there have been profound technical and organisational changes in pig production. Technical changes involve the increased professionalism with intensive use of micro-electronics in data recording, communication and processing, better reproductive technology and exploitation of gene technology. These changes have led to an increase in the initial investment demands. Organisational changes have been driven by increased competition; the whole production chain has increased in scale, areas such as Latin America, Southeast Asia and East Europe have been re-introduced to the world market, the North American industry has grown strongly, and international trade has been liberalised. Pig breeding is becoming more and more international. Pig breeders are therefore confronted with an increasing need to disseminate their genetic improvement to more end products to channel more funding back to the nucleus level and an increasing variation in the production conditions of their customers together with an increasing professionalism among them and steadily reduced margins (Knap, 1998). In hybrid breeding, standardised and highly efficient end products are bred, which are excluded from further breeding. For pigs these are hybrid fattening pigs. They are normally produced by breeding units with the respective multiplication systems, which can produce hundreds of thousands of hybrids if demand arises. In the past decades in the pig sector these hybrid-products have displaced conventional breeding units which produce less 286 DEVELOPMENT OF PIG GENE FLOW AFTER WORLD WAR II standardised, less efficient and lower numbers of fattening animals. Inevitably this has reduced genetic variation considerably. International pig breeding companies like PIC usually have a variety of exports. The first is the one-off purebred consignment targeted at breeders that want to replenish their herds with a few animals. This kind of transfer is suitable for locations both far and near, but the individual value and with it cost of the single animals is usually high because the company does not expect much repeat business. The second is the transfer of grandparent or great- grandparent stock for breeding programmes. This involves supply of complete herds of boars and gilts, which can then produce further breeding stock in the importing country. This often happens with the supervision or even ownership of the exporter. Again this type is also suitable for remote locations. Finally there is the regular transfer of hybrid parent stock. This is only suitable for short distances with low transport costs, since the parent stock must recoup investments within one breeding lifetime (Bichard and Dyson, 1980). As well as considering the type and distance of planned transfers, an important factor is the adaptation of the genetic stock in their hybrid programmes to the environmental conditions for which they are intended. For example, it may be inappropriate to introduce the Pietrain as a terminal crossing partner in Southern China because it can be too sensitive. Furthermore consumers may prefer fat to lean meat, and this requires adaptation. International breeding companies take account of this by establishing nucleus herds in the target regions or countries they wish to service and adapt the animals there. Apart from hybrid programmes transfer of breeding pigs may take place in other breeding systems. Rotational systems are in theory quite efficient but are, in practice, difficult to run properly. They have, in the past, mainly served as the first stage of improved breeding systems by allowing farmers to partly exploit crossbreeding advantages. Today they have largely been replaced. There are also important synthetic lines in use today. However, they are not generally used as new breeds in purebreeding systems, but instead contribute to modern hybrids for some markets. As an example, PIC uses the Leicoma, which was developed in the large pig production units in the former GDR to withstand the prevailing hard physical conditions, as one component of a hybrid.

3 DEVELOPMENT OF PIG GENE FLOW AFTER WORLD WAR II

3.1 Influence of technology and global mobility on the spread of breeding pigs and changes in breeding organisation Breeding pigs worldwide are now mainly distributed by a few international breeding companies. This is particularly true in the case of hybrid pigs. Although in the past developed countries contributed most to the dissemination of genetic material, nowadays there are strong breeding enterprises e.g. in Thailand, the Philippines and China. The mobility in the second half of the 20th century has increased tremendously compared to previous centuries. However, although technically transport of any magnitude and distance is possible, freight costs are the main restriction to transfers, but there are also veterinary restrictions, welfare considerations and exchange rates and inflation difficulties (Bichard and Dyson, 1980). Consequently almost always, airfreight of breeding pigs is at breeding weights of 100 kg or lighter to save freight costs. DEVELOPMENT OF PIG GENE FLOW AFTER WORLD WAR II 287

Apart from logistics for global mobility international breeding companies strongly utilise all biotechnological methods. However, still today, mainly live breeding pigs are transferred. Ova are unimportant because it has not been possible to store them. Reliable semen freezing was slow to develop, but is now being used to maintain improvement in the purebred lines, which are exported from the highly technical nucleus units in North America or Europe. Commercial production increasingly relies on AI, the semen coming from imported or locally bred boars. This mode of transfers enables also elaborate disease control, to prevent diseases being exported into the new countries. International breeding companies are operating under high pressure of competition. This leads to an increasing tendency towards monopolisation. This implies a danger of genetic loss if no safeguards are put in place. After 1945 national, regional and commercial pig breeding programmes began to develop. While governments or breed society-owned programmes were often pleased to have the opportunity to export a few breeding pigs, their primary focus was on their home markets. This contrasts with some of the commercial programmes, which actively sought export opportunities. They usually tried to retain control of their purebred lines and so the main saleable items were crossbred breeding stock. But because transport costs are high relative to the value of the stock, it is not usually economic to export large volumes of breeding females (except for short distances, e.g. by truck from the United Kingdom to Germany). Usually the company sent a small number of purebred stock (either to a daughter company or a licensed distributor) to establish multiplication herds. These herds then produced crossbred breeding animals for sale in the overseas territory. While the herdbook sector and some commercial farms in the receiving countries will continue to import small quantities of breeding stock with no restrictions on use, most exports from the West will be controlled. In the past, the breeding companies might have established breeding farms abroad, which were wholly owned or joint ventures with local partners. Today, because of the need to minimise their capital investment and the growing size and sophistication of the foreign commercial pig producers, they may incline to sell breeding stock under contract. The local producers agree, for example, not to sell breeding stock to third parties, to allow the breeding company to examine their record system, and to pay it a “genetic royalty” every time a new breeding animal produced within the multiplication unit is transferred to the breeding unit.

3.2 Main distributors of breeding pigs The main players in the distribution of breeding pigs are UK, Netherlands, Denmark, Sweden, Belgium, Hungary and USA. However, both absolute and relative export quantities are difficult to estimate, mainly because almost all companies are primarily focussed nationally. Their main role is to provide breeding stock for their home country. Generally commercial companies are more interested in selling relatively few high-priced pigs to an overseas customer than in establishing part- or wholly-owned businesses in those countries. The main profit of some international breeding companies comes from genetic royalties and this increases as its customers change to large . The amount (per parent gilt) is determined by what the profits would traditionally have been had (when the customer only purchased parent gilts) less the savings to the breeding company from not having to own or contract multiplication farms, fleets of trucks etc. The 288 DEVELOPMENT OF PIG GENE FLOW AFTER WORLD WAR II customer gains from not having to pay the full “market price” for a parent gilt but of course picks up some of the costs of doing the multiplication “in house”. In the following the main nations are being characterised briefly. The UK never had strong breed associations and (after boar licensing ceased 30 years ago) had no legal support, such as was the case in Germany for example. The British Ministry of Agriculture was heavily influenced by the new science of population genetics, which first flowered in USA but soon had a world-class centre in Edinburgh, and it left the industry free to develop new organisations and breeding plans. The Meat and Livestock Commission, funded by commercial pig producers, and the meat trade, put heavy emphasis on improvements within purebreds (building a chain of performance testing stations) and encouraged individual breeders to combine and market crossbred breeding pigs. Government research centres pioneered crossbreeding studies sooner than most other European countries. Additionally a strong forward-looking spirit helped, with conservative institutions that were much weaker than those in mainland Europe. Being an island made disease control easier so that pigs were more acceptable to importing countries’ veterinary authorities. The largest commercially owned British breeding company is PIC (now Sygen), which claims to supply 40% of the annual replacements in the USA, based upon purebreeding herds mainly imported from Europe since 1973. PIC produces between 2 and 3 million replacement gilt equivalents annually worldwide today. Equivalents are used since actual sales will be a mixture of parent gilts (used directly to produce slaughter progeny) grandparent gilts (used in customers‛ multiplication herds to produce parent gilts) and great-grandparent or nucleus-level gilts (one further level up the breeding pyramid). PIC weights each category by its expected lifetime output of parent gilts (thus grandparents might produce five selected daughters in each of 2.5 years - hence a factor of 12.5). NPD (Northern Pig Development Company) renamed National Pig Development Company was a family-owned business grown from the 1960s, which made important sales worldwide and to USA. It was purchased by PIC around 1995, having then an approximate size of one third of PIC. Cotswold Pig Development Company, with origins in the 1960s, was owned by the Nickerson family for many years. It reached a similar size to NPD but declined over the past 10 years and has been broken up and sold off. Newsham emerged in the 1980s as an offshoot of NPD and made significant sales to Mexico and USA. Both Cotswold and Newsham were acquired by JSR, a breeding company, which was built up by John S. Rymer from the 1960s. It is now the second largest UK company after PIC. Masterbreeders originated in the 1960s when it was owned by a UK feed company, which was part of Unilever. It conducted a lot of business in the Far East but eventually failed. The Netherlands has a large and nationally important pig industry and a well-developed, co-operative system with a strong trading tradition. Both herdbook and private (e.g. Unilever) improvement programmes were able to develop. For example, Dalland is a commercial company with significant worldwide exports that was originally part of the Unilever group but then absorbed (15-20 years ago) into the large Stamboek herdbook DEVELOPMENT OF PIG GENE FLOW AFTER WORLD WAR II 289 based company. The main commercial companies that play a role in the export market today are Topigs and Hyporc. Denmark started very early with a strong national programme (1900) based on co- operative organisations of farmer-owned meat exporting factories. But after 1945 they decided that exporting breeding pigs could damage their meat exports and subsequently banned pig exports. As a result, UK had to import its Landrace stock from Sweden as second-best in the early 1950s. But Denmark eventually realised its mistake in relying totally on its one breed (Danish Landrace) and imported British Large White and US Durocs and Hampshires in the 1960s and eventually started up breeding pig exports. Post 1970, the national programme of the farmer owned slaughterhouses re-entered the export business. However, Denmark had missed the opportunity to be a world leader. Sweden gained importance in the export business mainly due to its role in serving the market demand for its Landrace, as long as Denmark remained closed. Indirectly Sweden plays a role via the UK’s Landrace. Belgium possesses uniquely muscled breeds (of all species), and their Belgian Landrace and Pietrain were in demand when everyone wanted leaner pigs. Seghers, a long established company, is the only one to commercialise this trade and has been an important exporter for many years. However, the herdbook sector/ministry programme has also exported purebred stock worldwide (including to hot countries where these stress- susceptible pigs had high mortality and poor muscle quality). Hungary was the most important central European country in the trade with breeding pigs. One of their state-run organisations (Babolna) had the large-scale facilities of most of the central European countries, but also adopted good genetics and good marketing. So its pig breeding programme for large industrialised production systems (Tetra) was quite widely exported (as also its hybrid poultry). Tetra exported not only to other central European countries, but also to Southeast Asia pre-1985. In the USA, with its free-enterprise economy, and where quantitative genetics and statistics were first applied to improving crops (maize) and livestock (poultry), it would be expected to find genetically improved pigs and exports. New organisations were created: Farmers Hybrid and later De Kalb and Kleen Leen. But early improvement methods in corn involved creating and then crossing inbred lines. This was not as successful in pigs and held up their progress for years. The really interesting question is not why USA was influential in world gene flows in pigs, but why it did not dominate the world, and even allowed European companies to arrive in the 1970s and dominate US pig production today? The answer may lie in the different US carcass market (their pigs remained fat for years) and in the support the universities continued to give to the small-scale local purebred breeder and the show ring, rather than educating pig producers to put their trust in science-based breeding organisations. Only when large-scale agribusiness replaced the small family farm the demand changed to more efficient sows and leaner carcasses. Exports to Central and South America and to Southeast Asia reflected US aid to these regions, the large US pig population, and the existence of US breeding methodology in plant and poultry breeding - even at a time when US pig improvement methods were still very traditional. Both Farmers Hybrid and De Kalb have exported to Latin America and Southeast Asia, but not extensively to Europe. The size of De Kalb is mainly reflected by 290 DEVELOPMENT OF PIG GENE FLOW AFTER WORLD WAR II the large size of its home market, however it also established De Kalb companies abroad, for example in Japan. Additionally PIC operates from USA and Smithfield Swine Genetics plays a role, which uses stock derived from NPD in UK. Further players in the commercial exchange of genetic material are Canada with Geneticporc and Donaldson, France with Nucleus (French purebred breeders), Pen-ar-Lan and France Hybrides, Ireland with Hermitage, Spain with Bataille and Germany with the Bundeshybridzuchtprogramm (German Hybrid Pig Breeding Programme). In conclusion, the important source countries were those which controlled some unique genotypes (Large White, Scandinavian Landrace, Pietrain, Belgian Landrace, Duroc, Hampshire), and then those which emerged quickly from the old, conservative breed- society and government-dominated structures (UK, Netherlands, USA).

3.3 Main recipients of breeding pigs Central and South America imported their ideas and breeding stock initially from USA. US professors persuaded them that rotational crosses using Yorkshire, Duroc, and Hampshire were what they needed; and large numbers of boars went south, mainly from small breeders but also from the US breeding companies. But when corporate pig production started in the late 1970s, these southern countries were much more impressed by the intensive European management methods and bought European hybrid sows to populate their large new units, often with US-breed boars. Thus Mexico, Chile, Brazil, Peru, Argentina, Columbia, and Venezuela all bought from PIC though the European stock often came via US herds. Uruguay reports imports for their commercial hybrid pigs from both Brazil and Spain (Country report of Uruguay, 2003). Incidentally, Cuba bought North American breeds from Canada because they do not trade with USA. Some South American countries utilised muscular European breeds (Pietrain, Belgian Landrace, German Landrace) e.g. Brazil - in part because their German/Italian derived pig producers had contacts there. As in South America, US professors advised many countries in Southeast Asia (Japan, Thailand, Singapore, Taiwan, Philippines) to adapt their rotations and breeds. The US stock went into the purebred herd structure or else into a few large breeding programmes run by feed millers, brewers or others who were entering agribusiness. Corporate pig production was often started by corporate poultry producers who had probably successfully imported Western poultry genotypes. But there were often disasters because the imported breeding stock could not always resist the heat and disease burdens in the rather unsophisticated Asian units, and their carcasses did not always satisfy the very specific local demands. But once again, as production methods became more sophisticated, demand increased for European hybrid sows even if these were supplied via US-based breeding companies. For example: PIC (now Sygen) is a London-based public company, but its major business is in North and South America and it often supplied its Asian customers or joint ventures or subsidiaries from its North American breeding units. Australia should perhaps be included in the list of exporting countries, but until 10 or 15 years ago its impact was not significant. Its extremely strict health regulations kept out most imports so that its own pig herds have not benefited from the early European or US advances. The only export market it services is Asia, dependent mainly upon lower transport costs, good health of pigs and the persistence of Australian traders active in those countries. The Australian pig production industry is perhaps more dominated by large- DEVELOPMENT OF PIG GENE FLOW AFTER WORLD WAR II 291 scale agribusiness than any other, and these companies can only expand by exporting meat or pigs. Compared to South America and Asia, the influx of breeding pigs to Africa has been negligible. In South Africa strict veterinary regulations made imports almost impossible. PIC’s franchise there mainly had to use their existing breeds but apply Northern European methods, although UK pigs from NPD did get in. Apart from the main planned flows from Europe and North America into developing countries of the South, there is unplanned "leakage" or gene flow from primary customers to the whole industry in those and neighbouring countries. Often, purebred pigs were not sold, except under contracts, which prohibit or control purebreeding, but it has been impossible to keep control completely. Additionally commercial companies may establish joint ventures, daughter companies or associates in emerging markets. Important factors for such developments are availability of animals when needed, import health requirements and the assurance that the company can profit financially from the transactions. PIC supplies the pig breeding sector in new countries from its main bases in UK and USA. However Chile was responsible for supplying breeding pigs to Peru, while Argentina was supplied by Chile and Brazil. New Zealand received its breeding pigs from Australia and Canada sent boars to Australia (Table 1).

Table 1: Early PIC joint ventures, daughter companies and associates worldwide

Year Country Type Pigs generally derived from 1970 Canada Joint Venture UK 1972 Mexico Joint Venture UK via Canada 1973 Australia Daughter Company but semen sent out subsequently 1973 USA Daughter Company UK via Canada 1977 Brazil Joint Venture UK via France 1978 Chile Associate USA 1982 Japan Associate USA 1984 Korea Associate 1985 China - Hubei Associate 1985 New Zealand Associate, stock from Australia and Canada 1985 Peru Associate Chile 1988 Cuba Fixed term contract 1992 Argentina Joint Venture 1993 Colombia Associate 1993 Puerto Rica Associate The size of these ventures reflects the size of each national market, since each is largely confined to within its own country. But of course the competence of the individual 292 CURRENT STATUS AND ACTUAL TRENDS OF PIG GENE FLOW company has also been important. Thus, the Japanese company grew very slowly and the Chinese venture in Hubei failed to make much progress as it was more profitable to supply slaughter pigs to Hong Kong than to develop a nationwide business selling breeding stock. By contrast, the associate in Chile was extremely well managed and grew very rapidly to dominate the local industry and greatly increased the per capita consumption of pork. Similarly the Brazilian company grew quickly to a large size. After 1988 the gene flow of pigs became more complex. Pigs were moved when needed from whichever PIC-controlled herds had them available and healthy. This latter point is most important. Today the company has a large nucleus (purebreeding) herd in Southeast Saskatchewan (Canada). It was specifically established in a very isolated area away from people and other pigs to try to remain free of infections. Canada as a whole has a good international reputation for pig health. As a result, many international pig transfers originate from this Canadian herd which contains breeds or lines sourced originally from all round the world. In summary, in the first four decades after World War II all countries in the Americas were strongly influenced by breeds from the USA (Hampshire, Duroc and Yorkshire). The same thing occurred in Japan, Taiwan, Korea, Thailand, Singapore. Similarly, purebred stock from Britain went to the old commonwealth countries (Australia, New Zealand, South Africa, Kenya, Zimbabwe) using Large White and Swedish Landrace.

4 CURRENT STATUS AND ACTUAL TRENDS OF PIG GENE FLOW

4.1 Gene flow indicated by selected export and import data for live breeding pigs Compared to ruminants very little information on the current exchange of breeding pigs is available. For Europe, the Eurostat statistic database was accessible for purebred pigs. No other statistic databases were available. Therefore, publications and breeding associations needed to be consulted, but the report remains sketchy. Particularly if considered, that crossbred animals, which do not appear in the Eurostat statistic database, have gained increasing importance. Additionally, non-traditional countries for breeding pigs, such as Thailand, are gaining increased market shares. In commercial trade with breeding pigs the following countries compete with each other: Britain, USA, Canada, Belgium and Netherlands, Denmark and Hungary (Bichard and Dyson, 1980). Nowadays also France, Germany, Ireland and Spain play an important role, and Hungary has lost importance.

4.1.1 European Union

Commercial pig breeding companies which have established bases in second countries from which to supply third country customers are PIC (Sygen) from UK, Topigs and Hyporc from Netherlands. However, as the market for breeding stock has become more difficult within European countries in recent years, every national company has taken more interest in exports and have made opportunistic sales worldwide. The following data is compiled from the Eurostat database, maintained by the European Union and collected by the member states (e.g. through customs declaration). The EU refers to the EU-15 countries (Austria, Belgium, Denmark, Finland, France, Germany, CURRENT STATUS AND ACTUAL TRENDS OF PIG GENE FLOW 293

Greece, Ireland, Italy, Luxembourg, Portugal, Spain, Sweden, The Netherlands, United Kingdom), the new member states are not yet considered. Between 1992 and 2003 EU countries exported an annual mean of over 125,000 purebred breeding pigs. This figure masks however dynamic changes: from 1992 onwards exports constantly grew from about 54,000 to about 105,000 in 1997. Thereafter exports fell to less than 59,000 in 2001. In 2002 exports from EU countries sky-rocketed to almost 390,000 and fell to about 336,000 in 2003, which was in average almost five times the 1992 to 2001 average of about 78,000 (Figure 1).

Figure 1: EU exports of purebred breeding pigs (highlighting the Netherlands) by destination and share of extra-EU exports

400 extra-EU 40% intra-EU 14 exports from Netherlands 350 35%

extra-EU share 13

309 300 30%

250 25% 232

200 20% 376

323 15 0 15 %

10 0 10 10 %

number of animals in thousand animals of number 12 9 7 11 8 23 12 5 50 9 95 5% 80 80 70 77 73 46 48 54 54 11 6 8 776 0 124 3 0% 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Source: Eurostat (2004)

The Netherlands were responsible for the massive growth in exports of purebred breeding pigs in 2002 and 2003. Their market share in EU-exports rose from 10% in 2001 to 79% in 2002. In 2002 the Netherlands reported exports of purebred breeding pigs to Germany of almost 270,000 and in 2003 of almost 160,000 purebred breeding pigs to Spain and more than 40,000 purebred breeding pigs to Germany. The sudden jump in Dutch exports is probably a distortion; almost certainly these figures result from re-classifying animals destined for finishing and slaughter as “purebred breeding pigs”. However, they also reflect structural changes in the importing countries, particularly in Germany after the re- unification process. France almost continuously increased its exports from the early 1990s until 2003 while exports of Belgium-Luxembourg decreased strongly. In absolute numbers UK had a more or less constant export of breeding pigs (average 44,000 per year) in this period, though its share in EU-exports fell due to the massive increase from the Netherlands. 294 CURRENT STATUS AND ACTUAL TRENDS OF PIG GENE FLOW

From 1992 to 2001, UK was the main exporter of the EU countries with a mean share of 29% of all exports. It was followed by Belgium-Luxembourg (18%), France (16%), Spain (11%), Denmark (8%), the Netherlands (7%), Ireland (5%) and the other EU countries (6%). Only in 2002 Netherlands gained importance as described above. Most of the reported exports were to other EU countries (115,000 purebred breeding pigs). Only about 11,000 purebreds left the EU, 8.8% of the total exports. This share declined almost continuously from 16% in 1992 to 3.9% in 2003, interrupted by a peak in 1995 of almost 30%. The relative importance of purebred breeding pigs leaving EU countries is decreasing (Figure 1). After 1990 there was a considerable flow from EU countries to former communist East Europe as these countries sought to catch up, particularly in terms of carcass quality. These pigs were sourced both from small pedigree breeders (perhaps government backed schemes) and hybrid companies. The Eurostat data confirm the main trading partners between 1993 and 2003 to be countries that joined the EU in 2004 - Poland and the Czech Republic being the most significant (Figure 2), closely followed by others. 21% of extra-EU exports went to East and Southeast Asia: Thailand, South Korea, Japan and China are the largest trading partners, followed by The Philippines and Vietnam. The remainder went to Taiwan, Malaysia and Singapore and marginal shares to Hong Kong and North Korea. Some minor exports to Asia went to Kazakhstan, India and Bhutan. 16% of the extra-EU exports went to North America. Between 1992 and 2003 the share of the USA was about twice as big as the one of Canada. About 9% extra-EU exports went to Latin America, more than half to Brazil. Also Mexico made up for considerable numbers of breeding pigs exports from the EU especially in the first half of the 1990s. With some distance Venezuela, Colombia, Guadeloupe, Uruguay, Argentine and Martinique followed. Also some other Latin American countries imported relatively small number of breeding pigs from the EU. 2.7% went to Sub-Saharan Africa. Main trading partners were Cape Verde, South Africa, Nigeria and Kenya. Close to 1% of extra-EU exports went to the West Asia/Maghreb region, Turkey and Tunisia being the main partners. Although these countries traditionally do not have a large pork market, a demand for breeding pigs can be attributed to the flourishing tourism industry. Marginal numbers of purebred breeding pigs went to Oceania and unspecified destinations. CURRENT STATUS AND ACTUAL TRENDS OF PIG GENE FLOW 295

Figure 2: Annual EU purebred breeding pig export by destination 1992-2003

Poland 958 Czech Rep. 778 Romania 540 Slovakia 473 Hungary 385 CIS 328 Bulgaria 181 Canary, Ceuta 777 Croatia 392 Cyprus 66 Slovenia 47 Bosnia-Herz. 20 Albania 19 Serbia and Montenegro 18 For.JRep.Mac 12 Malta 5 Yugoslavia 0 Latvia 245 Lithuania 212 Estonia 16 Norw ay, Iceland 9 Sw itzerland, Liechtenstein 13 South Korea 398 Japan 386 China 385 Taiw an 75 North Korea 19 Hong Kong 13 Thailand 426 Philippines 322 Vietnam 176 Malaysia 62 Singapore 41 Kasakhstan 29 India 8 Bhutan 7 USA 1,127 N.

Am. AsiaCanada Europe 617 Brazil 531 Venezuela 60 Colombia 47 Uruguay 36 Argentina 34 Chile, Fr. Guiana, Peru, Falkland Is., Surinam 25 Mexico 214 Costa Rica 5 Guadeloupe 44 Martinique 24 Dominican R., Dominica 0 Cape Verde 104 Nigeria 50 Ghana 10 Giunea, S. Helena, Burk. Faso, Benin, I. Coast 9 South Africa 79 Gabon 13 Cameroon, Congo, S. Tome 5 Kenya 24

E. C. W. Carib. C. S. SCZimbabw SE e 4 E. N. S. E. Malaw i 3 Tunisia 43 Turkey 25 Bahrain 14 Saudi Arabia 12 W. Asia/ Maghreb Africa America Latin Georgia 11 Aust.Oceania 3 not specified 1 0 250 500 750 1,000 1,250 number of animals

Total intra-EU export 114,631 and extra-EU export 11,013

Source: Eurostat (2004) 296 CURRENT STATUS AND ACTUAL TRENDS OF PIG GENE FLOW

In the years 1992 to 2003 EU countries imported an average of 134,000 purebred breeding pigs annually (Figure 3). 1996 was the year with the highest imports (551,000), mainly through 450,000 purebred breeding pigs imported by the UK from Ireland. These may be a severe distortion. The UK probably only needed a total of 250,000 replacement breeding pigs in 1996, most of which would originate in UK, so it would not import 450,000 from Ireland. Perhaps these were weaner pigs brought into UK for finishing, or sent on into mainland Europe via UK. Other large transfers were reported by Germany in 1999 of almost 44,000 purebred breeding pigs from the UK, by Belgium in 2001 of almost 38,000 purebred breeding pigs from the Netherlands and by Spain in 2002 of almost 25,000 purebred breeding pigs from the Netherlands. For six EU countries (Belgium, France, Greece, Italy, Luxembourg and Spain) the Netherlands were the main source of pig genetic material and for three countries (Ireland, Denmark and Germany) the UK during the time period from 1992 to 2003. Exports from UK are mainly attributed to PIC. Information given in Figure 2 and Figure 3 is not fully compatible.

Figure 3: EU imports of purebred breeding pigs by destination and share of extra-EU imports 1992-2003

600 40% 1 intra-EU extra-EU 550 extra-EU share 35% 500

450 30%

400 25% 350

300 20% 550 250 15% 200 4 2 2 2 2 number of pigs in thousand in of pigs number 150 3 10% 3 100 0 1 134 128 126 138 140 5% 50 1 119 0 70 83 30 50 0 19 0% 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Source: Eurostat (2004)

The main origin of purebred breeding pigs is the EU. Only an average of 1,775 purebred breeding pigs annually (1.3%) were imported form extra-EU countries. 98% of these extra- EU imports originated from other European countries, 96% were from countries that accessed the EU in 2004, mainly Cyprus, Czech Republic and Hungary. Imports from other continents were marginal compared to gene flow within Europe: in 2000 176 purebred breeding pigs were imported from Argentine to France, between 1997 and 1999 a total of 110 purebred breeding pigs were imported from Canada to Italy and France, in 1992 the UK imported 12 purebred breeding pigs from the USA and between CURRENT STATUS AND ACTUAL TRENDS OF PIG GENE FLOW 297

1993 and 1995 the UK imported 16 purebred breeding pigs from Australia and New Zealand. Other trading partners for imports to the EU were not mentioned between 1992 and 2003. Although the absolute numbers are small, it can be assumed that these were high-value nucleus animals and the gene flow caused by these imports is important. However, due to lack of breed information and follow-up data, this is only speculative. For the shares of imports of pig breeding animals according to regions of origin refer to Table 2.

Table 2: Mean annual EU imports of purebred breeding pigs 1992-2003 by origin

Number of pigs EU 19,427 EU accession countries 2004 Cyprus 1,317 Czech Rep. 338 Hungary 50 Other European countries Switzerland 24 Norway 16 Extra-European countries Argentina 15 Canada 12 USA 1 Australia 1 New Zealand 1 Total 21,202

Source: Eurostat (2004)

The three main exporting countries (Netherlands, France, Denmark) had import-export- ratios of 0.16, 0.32 and 0.01 respectively. Also Ireland, Austria, Sweden and Finland were net-exporters. The three largest importing countries are UK, Luxembourg and Germany, with import-export-ratios of 1.8, 1.9 and 14.7 respectively. Other net-importers were the countries in the southern EU namely Spain, Italy, Greece and Portugal. For a comparison of absolute numbers refer to Annex 9.1 A 50. It must however be considered that there may be distortions due to misclassifying breeding pigs in the statistical bases. The European Union up to 2003, with its 15 member states, has a high trade volume with purebred breeding pigs. However only very few transactions of purebred breeding pigs with outside nations can be observed. This is true for exports and more so for imports. Of the few extra-EU transactions the main share lies with other European countries. The interpretation of the Eurostat data is however difficult. There is no information on breeds, or the organisations conducting the trade. Large transfers normally indicate also a large gene flow. International breeding companies, which import and export purebred breeding pigs to their own daughter companies usually induce high gene flow, whereas single transactions by individual breeders usually result in low gene flow. However, in certain cases single transactions may be successful and so cause high gene flow. Follow-up information is rarely available in order to verify the extent of gene flow. 298 CURRENT STATUS AND ACTUAL TRENDS OF PIG GENE FLOW

It is further unclear how records reflect the true volumes and whether transfers as part of development projects are also recorded. Often figures are difficult to explain and may be distorted, possibly due to problems with categorising pigs into purebred breeding pigs. Hybrid gilts are not recognised as breeding pigs, so that they have to fall under other pigs. Also there may also often be a strong incentive to falsify the category in order to avoid paying taxes or in order to secure a subsidy.

4.1.2 North America

The information available for North America is limited, and generalisation or representativity cannot be achieved. In Canada, between 1996 and 2003 the Canadian Swine Breeders’ Association reported mean annual exports of 473 breeding animals. It peaked at 1,260 breeding pigs in 1997 and had its minimum in the following year with only 78 exported breeding animals. Figure 4 illustrates this development. The figures quoted are most likely smaller than in reality since not all breeding pigs are registered through the Association, e.g. PIC’s main nucleus is located in Canada and these numbers would not be included here. Figure 4: Exports of breeding pigs from Canada

1,400 1,200 1,000 800 600 1,260 400 880

number of animals number of 571 200 200 226 0 78 97 1996 1997 1998 1999* 2000 2001 2002 2003 *data not available

Source: own elaboration, data from Canadian Swine Breeders’ Association (2004)

4.1.3 Asia

Quantitative data available from Asia is scarce and the countries and figures presented here are not representative. With quantitative data on China missing altogether, general conclusions may not be drawn. A few examples however illustrate some aspects of Asian pig trade. Thailand imported breeding material for its pig sector almost exclusively from industrialised countries, mostly Large White and Landrace breeds. Pietrains were imported from Belgium and Hampshires from France. British breeding companies, imported during and after the Foot-and-Mouth disease outbreak from Poland and Australia to Thailand. Table 3 gives some indication of the extent of gene flow to Thailand. CURRENT STATUS AND ACTUAL TRENDS OF PIG GENE FLOW 299

Table 3: Pig breeding material imports to Thailand

1998 1999 2000 2001 2002 2003 Semen 850 1,330 608 24,700 930 3,000 doses Breeding Male 430 313 400 750 164 animals Female 1,204 1,070 450 990 268

Source: Department of Livestock Development Thailand (2004)

In 1983, a total of 2,397 breeding pigs were imported to Singapore. They originated from UK (63%), USA (16%), Australia (15%), Canada (5%) and New Zealand (1%). Breed composition was 64% Duroc, 21% Yorkshire, 13% Welsh and 3% Landrace (Quek et al., 1984). It cannot be concluded from the present figures whether the 1983 imports are representative for the decade. With gifts from China, France undertook crossbreeding experiments in 1979 based on nine imported pigs of three breeds. The trials carried out are considered the first performance tests of Asian pig breeds under modern husbandry conditions. Genetic material from these trials spread to other European countries. In 1989, 140 Chinese breeding pigs were imported for experimental purposes to the USA after having fulfilled the required quarantine regulations for experimental crossbreeding purpose. Similar imports of 32 breeding pigs were organised in 1985 to the UK. Also countries like Albania, UK, Japan, Korea and Thailand have imported Chinese pig breeds (McLaren, 1990; Wood, 1990).

4.2 Gene flow of pigs indicated by the introduction of foreign genetic material in existing breeds The formation of commercial pig breeds in Europe and North America depended heavily on the introduction of foreign genetic material. The Large White, the Yorkshire and Berkshire in England were heavily influenced by imports from South China, and soon after their establishment they replaced indigenous breeds. In the late 20th century the Taihu, Fengjing, Minzhu and Meishan Asian breeds were of interest in European and US pig breeding not only for their prolificacy but also for meat quality (Wood, 1990), but their influence has been slight. In a global survey of new pig breeds introduced to major pig- keeping countries during the 1970s, Landraces from different origins and Pietrain were among the most common. (Sutherland et al., 1985). More details are given in Annex 9.1 A 51. Most interesting to European and US breeders has been the hyperprolificacy of some Asian breeds. Litter size of the purebred Meishan breed is considerably larger than those of European breeds and annual sow output may be 50% higher. Also of interest is the precocity of some Asian breeds. It gained importance as little progress was being made in breeding for fertility traits in European breeds due to the low heritabilities. It implied potential to increase economic performance by crossbreeding selected Asian breeds with European breeds. Additionally exploitation of hybrid vigour was expected. Yet comparisons of Chinese-Western-crosses with commercial crossbreds showed little economic advantage. More recent suggestions only refer to transferring relevant genes into commercial genotypes (Ellis, 1992; McLaren, 1990). 300 CURRENT STATUS AND ACTUAL TRENDS OF PIG GENE FLOW

The other way round, Asian countries have imported pigs of European origin for crossbreeding. McLaren (1990) reports imports of Middle Yorkshire, Berkshire and Kmirov breeds to China in the 1950s. In the 1960s English Large White, Soviet Large White and Landrace were imported. From the 1980s there are reports on the imports of Landrace, Duroc and to a lesser extent of Hampshire from Canada. Imports of Western meat-lard breeds to China and crossing with Chinese breeds led to the formation of 29 new lines, which have made a substantial contribution to meet growing demands for pork (Jing et al., 2001). It was found that with increasing shares of Western gene contribution daily weight gains, eye muscle area and lean meat percentage increased while litter size decreased. Bichard (personal communication, 2005) reports that many recent imports into China have had limited impact for various reasons. First, Chinese importers did not build efficient multiplication organisations and e.g. used the imported stock to supply slaughter markets because of attractive markets for lean carcasses or because they could not find large-scale customers to build new units. Secondly, disease problems due to poor understanding of disease control methods led to infection from surrounding local stock. Thirdly, China only opened up in the 1990s to wholly-owned ventures of Westerners. After that e.g. PIC has established a successful breeding nucleus near Shanghai, which is now building a substantial multiplication pyramid as the new commercialism allows modern units to be established which can properly manage Western hybrids. Developing countries attempted to overcome productivity bottlenecks of pigs using crossbreeding schemes to exploit hybrid vigour (Omeke, 1989; Pathiraja, 1986). It was stated that low productivity does not merely derive from low management intensities but also from the low genetic potential of animals and the lack of crossbreeding schemes. Breed comparisons of indigenous Desi, exotic Landrace and Yorkshire and their halfbreds were subject to analysis in Indian state farms and may serve as an example at this stage: The purebred exotics led the ranking, followed by the half-breds while the indigenous breed was last. These results were similar in comparable studies (Sharma et al., 1998). Under village conditions in India Landrace and its cross with Desi was superior to Desi purebreds but inferior to Large White and its cross with Desi (Sharma et al., 1992). The ranking strongly depends on the production conditions. The genotype x environment interaction is larger the more different the production conditions are. The situation in Vietnam has been looked at separately in the case study: Impact of the use of exotic compared to local pig breeds on socio-economic development and biodiversity in Vietnam. Uruguay’s pig stock consists principally of commercial hybrid pigs with Large White, Landrace, Belgian Landrace and Pietrain, but also purebred exotic lines exist. A new breed called Pampa has been developed from Poland, China and Criollo, which is locally adapted to the extensive systems in the east of the country. The massive utilisation of imported hybrids is becoming a threat to local pig breeds. Details on the state of the pig genetic resources in Uruguay can be found in Annex 9.1 A 52. It was concluded that genotypes from temperate regions and their crosses face severe problems when confronted with management conditions in developing countries resulting in poor reproductive performance. Furthermore to avoid loss of heterosis in consecutive generations continual inflow of exotic germplasm was necessary. It cast doubts on the success of small-scale crossbreeding schemes. Without institutional, organisational and technical support these crossbreeding schemes were doomed to fail (Agbagha et al., 2001). CURRENT STATUS AND ACTUAL TRENDS OF PIG GENE FLOW 301

In some cases breeding methods including exotic breeds and established under smallholder conditions were neglected after the projects concluded. Without further access to exotic breeds, the impact of isolated crossbreeding on the local populations declined from generation to generation. However, also isolated projects with release of Duroc or Hampshire boars to smallholder conditions have left long time visual traces in local populations, as can be observed especially in least developed Latin-American countries. Applying different breeding methods for pigs in tropical conditions has limitations. We find up to now only a few isolated cases of selection efforts within local breeds and purebreeding of exotic tropical breeds. The main emphasis of breeding in the tropics is in discontinuous crossings and purebreeding of European-American breeds, especially in the commercial sector (Horst, 1994). In commercial pig production, as opposed to ruminant livestock, the environment is normally shaped according to the needs of the pigs. The adaptability problems are therefore in most cases reduced. Also commercially owned breeding enterprises often have regional structures in which breeding pigs are adapted to the local conditions. Although the main pig gene flow is expected to take place in commercial production, there is a continuous trend to replace local breeds with exotics. The degree of replacement varies. It is almost complete e.g. in Chile, Taiwan, Mexico, Spain (only a specialist outdoor section for top quality ham still uses Iberian breeds) and all the central European countries, which have nearly lost most of their breeds. In Denmark the Danish Landrace was universally used purebred before the 1970s; now most sows are Large White x Danish Landrace and the boars are based on Duroc and Hampshire. Much less replacement took place in Belgium, where Pietrain and Belgian Landrace still dominate, China or Brazil. In the latter two countries large agribusiness units use imported stock but small peasant farms still rely on local breeds though these have often been influenced by infiltration by boars or semen of Western breeds.

4.3 Pig Gene flow indicated by Country Report excerpts The excerpts on pig gene flow show generally high mention of pig imports (Table 4). These imports generally contain high-yielding breeds, mainly Large White, Landrace and Duroc. However, the extent of breed replacement is at the same time reported as low, except for Southern and Eastern Asia. This may be due to lack of information in the excerpts but it may also be caused by the difference in the production system into which these breeds are imported. Imported pigs are mainly used in intensive production systems located in more suitable regions with market access. In these cases there may be little interference with the local breeds kept in extensive systems. However the excerpts list no disadvantages to the exotic breeds, which is unlikely even in suitable production systems, so the excerpts may not give the true picture. 302 CURRENT STATUS AND ACTUAL TRENDS OF PIG GENE FLOW

Table 4: Country Report excerpts on gene flow of pigs

Breed Extent of gene Main breeds Breed Disadvantages Country report excerpts flow in and out of imported replace- of exotic breeds containing information the region ment In Out Africa low 0 Large White low Not named Burundi, Mali, Uganda, Zambia eastern and medium 0 - low Large White low Not named Belarus, Poland, Russian southern Federation, Serbia and Europe Montenegro, Slovenia Latin low 0 Landrace, low Not named El Salvador America Yorkshire, Duroc North high Low Yorkshire, low Not named Canada, United States of America Landrace, Duroc, America composites Oceania very high 0 Landrace, Large 0 Not named Palau, Samoa White Southern high - 0 - low Large White, medium Not named Bhutan, Cambodia, China, and Eastern very high Landrace, Duroc Indonesia, Japan, Asia Kyrgyzstan, Malaysia, Myanmar, Nepal, Philippines, Sri Lanka, Vietnam western and high - Medium Yorkshire, Large low Not named Denmark, Finland, northern very high White Iceland, Latvia, Europe Netherlands, Sweden, United Kingdom western low 0 Large White low Not named Armenia Asia

Narrative extracts from single country reports are summarised in Annex 9.1 A 53. As can be depicted from Annex 9.1 A 52, increasing worldwide spread of purebreds can be observed only for Landrace and Large White. However, the example of Japan (Annex 9.1 A 53) shows very clearly that, in the last 20 years, the share of purebreds has become insignificant, whereas “others, hybrids, crosses” contribute about 100% of the sows and 40% of the boars. This development can be observed in other countries at a slower speed and unfortunately can hardly be traced nor quantified. REFERENCES GLOBAL GENE FLOW OF PIGS 303

5 REFERENCES GLOBAL GENE FLOW OF PIGS

Agbagha, F.M., F.U. Ezema, and B.C.O. Omeke. 2001. Studies of management effects on fertility of purebred and crossbred exotic gilts in two breeding farms at Nsukka, Nigeria. Nigerian Journal of Animal Production 28:20-25. Bichard, M. and A. Dyson. 1980. Exporting British Breeding Pigs. RASE/ADAS/NFU Conference on "Pig Marketing - the Challenge of the 80s", Stoneleigh, 18 November 1980. Bichard, M. 2005. Personal communication. Canadian Swine Breeders’ Association. 2004. http://www.canswine.ca/. (05.07.2004) Comberg, G. 1978. Schweinezucht. Ulmer Verlag, Stuttgart, Germany. Country report of Uruguay for FAO’s State of the World’s Animal Genetic Resources. 2003. Ministerio de Ganaderia, Agricultura y Pesca. Department of Livestock Development Thailand. 2004. Unpublished. Ellis, M. 1992. Exotic breeds. The commercial potential of Chinese pig breeds. Meat Focus International 1:84-86. Epstein, H. and M. Bichard. 1984. Pig. In: Mason, I.L. (ed.) Evolution of domesticated animals. Longman, London and New York. Eurostat. 2004. EUROSTAT-COMEX Databank. European Union, Luxemburg. FAO. 2004. Global Livestock Production and Health Atlas (GLiPHA). 1[1]. 2004. FAO, Rome, Italy. Glodek, P. and M. Bichard. 1994. Analysis of the structure of pig breeding in EU countries with possible lessons for central and eastern Europe. European Association for Animal Production Task Force on Livestock Production in Eastern Europe. Workshop "East and Central Europe Livestock - Selp help in Livestock Production". January 1994, Berlin, Germany. Horst, P. 1994. Zuchtstrategien für tropische Standorte. In: Kräußlich, H. Tierzüchtungslehre, 4. Auflage, Ulmer Verlag, Stuttgart, Germany. Jing, R.B., C.Y. Song, Z.H. Liu, and Y. Tao. 2001. Breeding of new swine lines (female parent lines) in China in the past ten years. Pig News and Information 22:55N-58N. Knap, P.W. 1998. Internationalisation of pig breeding companies. 6th World Congress of Genetics Applied to Livestock Production, 12-16 January 1998, Armidale, Autralia. 26:143. Kräußlich, H. and G. Brem. 1997. Tierzucht und Allgemeine Landwirtschaftslehre für Mediziner. Enke, Stuttgart. Larson, G., K. Dobney, U. Albarella, M. Fang, E. Matisoo-Smith, J. Robins, S. Lowden, H. Finlayson, T. Brand, E. Willerslev, P. Rowley-Conwy, L. Andersson, and A. Cooper. 2005. Worldwide Phylogeography of Wild Boar Reveals Multiple Centers of Pig Domestication. Science 307:1618-1621. 304 REFERENCES GLOBAL GENE FLOW OF PIGS

McLaren, D.G. 1990. Potential of Chinese pig breeds to improve pork production efficiency in the USA. Animal Breeding Abstracts 58:347-369. Omeke, B.C.O. 1989. Field performance of exotic pigs and their crosses in tsetse-infested area of Nigeria. Bulletin of animal health and production in Africa 37:79-83. Pathiraja, N. 1986. Improvement of pig-meat production in developing countries. 1. Exploitation of hybrid vigour (heterosis). World Animal Review 60:18-25. Quek, K.L., C.F. Chang, C.E. Low, and T.H. Lee. 1984. Observations of imported breeding pigs during on-farm isolation. Singapore Veterinary Journal 8-9:15-19. Sharma, B.D., S.K. Singh, and A.A. Devi. 1998. Study on genetic and nongenetic factors affecting daily weight gain in exotic, desi and their halfbred piglets. National Academy Science Letters 21:324-329. Sharma, B.D., S.K. Singh, C.B. Dubey, and H.R. Mishra. 1992. A comparative study on growth pattern of exotic and desi pigs, and their halfbred under farm and village conditions of rearing. Indian Journal of Animal Science 62:378-380. Sutherland, R.A., A.J. Webb, and J.W.B. King. 1985. A survey of world pig breeds and comparisons. Animal Breeding Abstracts 53:1-22. Wood, D.R. 1990. The Meishan importation 1987. Pig Veterinary Journal 24:137-139. Gene Flow in Animal Genetic Resources. A study on Status, Impact and Trends

Editors: Anne Valle Zárate, Katinka Musavaya and Cornelia Schäfer

Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, Germany

Annex 9.6

UNIVERSITY OF AGRICULTURAL HOHENHEIM RESEARCH Institute of Animal ORGANIZATION Production in the The Volcani Center Tropics and Subtropics Institute of Animal Science Germany Israel

The Worldwide Gene Flow of the Improved Awassi and Assaf Sheep Breeds from Israel

Tobias Rummel, Anne Valle Zárate, Elisha Gootwine

VERLAG ULRICH E. GRAUER ‚ Beuren ‚ Stuttgart ‚ 2005 Tobias Rummel, Anne Valle Zárate, Elisha Gootwine: The Worldwide Gene Flow of the Improved Awassi and Assaf Breeds of Sheep from Israel.

VERLAG GRAUER, Beuren, Stuttgart, 2005.

ISBN 3-86186-497-5

© 2005 Institut für Tierproduktion in den Tropen und Subtropen Universität Hohenheim (480a), 70593 Stuttgart, Deutschland E-mail: [email protected]

All rights reserved.

Printed in Germany. Druck: F. u. T. Müllerbader GmbH Forststr. 18, 70794 Filderstadt, Deutschland

VERLAG ULRICH E. GRAUER Linsenhofer Str. 44, 72660 Beuren, Germany Tel. +49 (0)7025 842140, Fax +49 (0)7025 842499 Internet: http://www.grauer.de/, E-Mail: [email protected]

This case study is an independent part of the gene flow study implemented by the Institute of Animal Production in the Tropics and Subtropics of the University of Hohenheim. The Federal Ministry for Economic Cooperation and Development (BMZ) and the German Technical Cooperation (GTZ) acted as commissioner and project executing agency. The Food and Agriculture Organisation (FAO) acted as a support agency. An advisory panel composed of international scientists, representatives of donor and development agencies, the private sector and NGOs closely accompanied the study. The co-funding by the Deutsche Forschungsgemeinschaft (DFG) is gratefully acknowledged, as is the support of all sponsors.

ACKNOWLEDGEMENTS A variety of people from different countries and institutions contributed to accomplish this case study by providing helpful information and data on the gene flow of the Awassi and Assaf breed. These are Jock Allison from Abacus Biotech in New Zealand, Calle Escobar from the National Agricultural University La Molina in Peru, Yosef Carrasso from the Ministry of Agriculture and Rural Development in Bet Dagan, Israel, and Ebru Emsen from the Department of Animal Science at the Ataturk University in Turkey. Our thanks go also to Irina Florescu from the Ministry of Agriculture in Romania, Phillip Grand from Cowra Cheese in Australia and Sandor Kukovics from the Research Institute for Animal Production and Nutrition in Hungary. Finally, we wish to thank Nurlan Malmakov from the Research Institute of Sheep breeding in Kazakhstan, Ian McDougall (UK), Chanda Nimbkar from the Nimbkar Agricultural Research Institute in Maharashtra, India, Ch. Papachristoforou and C. Christofides from the Ministry of Agriculture of Cyprus, Daniel Roldao (Sociedade Agricola da Herdade do Matinho, Lda., Castelo de Vide, Portugal), and Makros Tibbo from the National Livestock Research Institute in Ethiopia. Not to forget Awassi Rt. in Hungary. In Israel, we are grateful to the devoted team of the Ein Harod flock that is engaged with the breeding of the Improved Awassi sheep for decades and that provided much of data for this study. We gratefully acknowledge contributions from all before mentioned institutions and persons to this study.

TABLE OF CONTENTS 311

TABLE OF CONTENTS

List of tables 313

List of figures 313

Terms and Abbreviations 314

Executive Summary 315

1 The Improved Awassi breed in its indigenous environment 316 1.1 Origin, domestication and distribution of the unimproved Awassi 316 1.2 Gene flow in the past of unimproved Awassi in traditional pastoral farming systems of Bedouins 317 1.3 Gene flow and breeding activities during the formation of the Improved Awassi breed 317 1.4 Gene flow and breeding activities to form the Assaf breed 318 1.5 Gene flow and breeding activities to form the Afec Awassi and Afec Assaf lines 318 1.6 Additional crossbreeding with the Improved Awassi and Assaf in Israel 318 1.7 Spread of the Improved Awassi and its crosses in Israel and the Palestinian Territories to intensive and extensive farming systems 319 1.8 Description and performance of the unimproved Awassi, Improved Awassi and Assaf breeds 320 1.9 Description and performance of the Afec Awassi and Afec Assaf lines 322 1.10 Conclusions 323

2 Worldwide gene flow of the Improved Awassi sheep from Israel 324 2.1 Middle East 329 2.1.1 Jordan 329 2.1.2 Iran 330 2.1.3 Abu Dhabi 330 2.1.4 Turkey 331 2.2 Mediterranean Region and Western Europe 331 2.2.1 Portugal 332 2.2.2 Spain 333 2.2.3 Italy 336 2.2.4 Cyprus 336 2.2.5 United Kingdom 337 312 TABLE OF CONTENTS

2.3 Eastern Europe and Central Asia 337 2.3.1 Hungary 338 2.3.2 Romania 339 2.3.3 Former Yugoslavia 340 2.3.4 Kazakhstan 340 2.4 Tropical countries 341 2.4.1 Ethiopia 341 2.4.2 India 343 2.4.3 Peru 345 2.5 Australia and New Zealand 346 2.6 Conclusions 349

3 References 351

4 Contact Addresses 356

LIST OF TABLES 313

LIST OF TABLES Table 1: Performance of unimproved, Improved Awassi and Assaf sheep in Israel 322 Table 2: Gene flow of Improved Awassi and Assaf breeding material from Israel 326 Table 3: Performance of Improved Awassi, Baluchi and Shal breed and their crossbreds in Iran 330 Table 4: Performance of Improved Awassi, Redkaraman and Tushin breed of sheep in Turkey 331 Table 5: Performance of the Malpica (Awassi x Mancha, Talavera, Churro and Castilian) breed in Spain 334 Table 6: Performance of Cyprus fat tailed and Chios sheep and their crosses with Improved Awassi in Cyprus on station 337 Table 7: Performance of Merino de Palas, Tigaie, Turcana and Improved Awassi in Romania 339 Table 8: Milk Production of Improved Awassi, Improved Awassi x Kazak Fine Wool and pure Karakul sheep on station in Kazakhstan 341 Table 9: Performance of the Improved Awassi and its crossbreds with local Menz sheep in comparison to pure Menz sheep in Ethiopia 343

LIST OF FIGURES Figure 1: Historical overview on Awassi and Assaf breeding in Israel 319 Figure 2: Improved Awassi ewes 321 Figure 3: Assaf stud ram 321 Figure 4: World wide gene flow of the Improved Awassi and Assaf breeds of sheep from Israel 324 314 TERMS AND ABBREVIATIONS

TERMS AND ABBREVIATIONS AI Artificial insemination ASBLACK ¾ Assaf, ¼ Barbados Blackbelly ASCEGA Association of Assaf Spanish breeders iAw Improved Awassi BB homozygous for the Booroola gene BC Basque Country B+ heterozygous for the Booroola gene CDR-AID US-Israel Cooperative Development Research Program C-L Castila-León C-M Castilla-La Mancha and Madrid CN heterozygous genotype; C=Callipyge gene; N=normal gene DANIDA Danish International Development Agency FecB Booroola gene ILRI International Livestock Research Institute KFW Kazak Fine Wool breed MASHAV Jordanian Ministry of Agriculture and Israel’s Centre for International Cooperation MV Maedi-visna MY Milk yield NARI Nimbkar Agricultural Research Institute Nav Navarre NSW New South Wales RIRDC Rural Industries Research and Development Cooperation UNALM Universidad Nacional Agraria La Molina UNDP/FAO United Nations Development Programme, Food and Agriculture Organization of the United Nations EXECUTIVE SUMMARY 315

EXECUTIVE SUMMARY In Israel, within-breed selection in the unimproved sheep started at the beginning of the 20th century by Jewish sheep breeders, resulting in the formation of the Improved Awassi strain. The Improved Awassi differs from the unimproved Awassi mainly by its remarkable high milk production – about 550 vs. 70 litres per lactation, respectively. To improve its prolificacy, the Improved Awassi was crossed in the 1960s with the East Friesian Milk sheep that was imported from Germany, resulting in the formation of the "Assaf" with improved prolificacy of about 0.4 lambs born per lambing over the Improved Awassi. Today, the Assaf has nearly replaced the Improved Awassi in Israel’s intensive dairy sheep sector. Recently, the Booroola gene, a major gene coding high prolificacy, was introgressed by crossbreeding to the Improved Awassi and the Assaf resulting in two new strains, the Afec Awassi and the Afec Assaf, respectively, both with a prolificacy of about 2.0-2.4 lambs born per ewe lambing. The gene flow of the Improved Awassi breed of sheep from Israel started in 1965. Since then, 28 transfers to 15 different countries are documented, with summarized monetary transfers over $2 million. Totally, 5,433 lambs, 1,100 doses of frozen semen and 143 embryos of the Improved Awassi have been exported from Israel. While only 9 of these 28 transfers have been part of development aid projects, the majority represent commercial transfers between private sheep farmers or government institutions and Kibutzim in Israel. Even though the biggest part of Improved Awassi breeding material, 2,944 lambs and 1,000 doses of semen, were transferred to Eastern Europe and Central Asia (Bulgaria, Former Yugoslavia, Hungary, Romania and Kazakhstan), the highest numbers of pure Improved Awassi and crossbreds with indigenous sheep are today in Spain (150,000-200,000 sheep) and Western Australia (100,000 sheep). The transfers to the tropical countries Burma, Ethiopia and India were part of development projects, but with limited success. In the Middle East, a total of 1,113 Improved Awassi lambs have been exported to Jordan, Iran, Abu Dhabi and to Turkey, where most of them were used to improve local stocks of unimproved Awassi. The following unofficial transfers to secondary destinations cannot be traced accurately. The gene flow of the Assaf started in 1977. 687 lambs, 11,354 doses of semen and 260 embryos, for a total price of $333,040, have been exported from Israel in 10 transfers to 7 different countries. Only 2 transfers were part of development aid, while the majority were commercial imports. The main stream of the gene flow of the Assaf breed of sheep went to the Iberian Peninsular. Portugal imported 6,854 doses of frozen semen and 260 embryos in the years 1991/1992, while Spain imported 430 lambs and 4,000 doses of semen between 1977 and 1993. In these countries the Assaf is today the dominating dairy sheep breed, numbering over 1.2 million pure- and crossbreds. Besides those two countries, the Assaf has been transferred to Peru, Jordan and Abu Dhabi, but with much less effect on the sheep production sector. From Portugal, Assaf breeding material was exported to the United Kingdom and Italy. No Assaf breeding material has been transferred to Eastern Europe, Central Asia, Australia or New Zealand. 316 THE IMPROVED AWASSI BREED IN ITS INDIGENOUS ENVIRONMENT

1 THE IMPROVED AWASSI BREED IN ITS INDIGENOUS ENVIRONMENT The following case study describes the worldwide gene flow of the Improved Awassi breed of sheep and of the Assaf breed that was developed from the Improved Awassi. While the native or unimproved Awassi sheep is the indigenous breed in different countries of the Middle East, the Improved Awassi sheep and the Assaf breed that developed from it are the achievement of an intensive breeding process that was started in 1932 by Jewish sheep breeders in Israel. This process is still going on today. According to the definition of the Israeli Sheep Breeders’ Association, only sheep which are either registered in the national flock book of the Improved Awassi, or the pure offspring from such sheep are called “Improved Awassi”. Today, the average milk production of ewes registered in the flock book is 506 kg in 214 days. Because of their superior milk yield, the Improved Awassi and Assaf of Israel have played a significant role in the world wide gene flow of sheep, in both numbers of breeding animals and material transferred for the last 40 years. Though there have been several transfers of unimproved or partly improved Awassi from other Middle East countries like Syria, Lebanon and Turkey, the following article is limited to lines of improved Awassi of Israel, because of their dominant role in the world wide gene flow of the Awassi breed, the short time frame those transfers were conducted and the detailed documentation available. This case study considers the history of the breed, its modification through crosses and its world wide distribution. The Awassi breed has been extensively described in the literature, in case of Israel especially by Hirsch (1933), Finci (1957), Epstein (1985), Gootwine and Goot (1996), Gootwine and Pollot (2000a) and Pollot and Gootwine (2001).

1.1 Origin, domestication and distribution of the unimproved Awassi The Awassi is the most numerous and widespread breed of sheep in southwest Asia (Epstein, 1985) and has evolved in the nomadic and transhumant production systems traditionally practiced in the Middle East region since biblical times (Epstein, 1971). It is assumed that the Awassi developed its characteristic fat-tail during over 5,000 years of domestication, which started in Mesopotamia, and that its parent stock were local breeds of thin-tailed wild sheep. Epstein (1985) remarks that heavy fat tails severely impede movement in the wilderness and can only develop in sheep protected by man. Because pigs are not consumed by Moslems or Jews, sheep were the main source of animal fat in the Middle Eastern region, supporting the selection of fat tail sheep where the fat is concentrated at the tail. Archaeological fragments and mosaics from Ur, dated back to 3,000 B.C., show horned sheep with a characteristic fat tail, indicating that the fat tail is an ancient product of domestication. Comparing the appearance of sheep on these ancient fragments with unimproved Awassi sheep reared today, it is obvious that the unimproved Awassi still represents the prototype of many related fat tailed sheep breeds in the Near and Middle East and can be considered as one of the oldest domesticated breeds of sheep worldwide. Its name is supposed to be derived from the Bedouin tribe “Awass”, living in the Euphrat region (Finci, 1957). Today, the Awassi is the dominant breed in Iraq, the most important sheep in the Syrian Arab Republic and the only indigenous breed of sheep in Lebanon, Jordan and Israel. It is also found in large numbers in Saudi Arabia and Turkey. THE IMPROVED AWASSI BREED IN ITS INDIGENOUS ENVIRONMENT 317

1.2 Gene flow in the past of unimproved Awassi in traditional pastoral farming systems of Bedouins The traditional sheep farming systems in the semiarid regions of the Middle East have been nomadic and in some regions transhumant, where animals and shepherds move during the entire year. Nomadic systems produce wider gene flows than sedentary farming systems, since animals are moved over long distances following seasonal pastures and water. Sheep keepers sell slaughtering and breeding animals on their movements to surrounding villages or nomadic tribes from other regions. For that reason the unimproved Awassi spread very soon after its domestication over a large geographic area, ranging from Turkey in the North and West to the Arab peninsula in the South and Iran in the East with only marginal differences in phenotype. Since the beginning of the 20th century this movement of the Bedouin tribes has been more and more restricted and is nowadays limited by political borders, forcing the tribes to change to a semi nomadic production system with less animal movement. Today, gene flow of Awassi sheep across national borders through nomadic Bedouin tribes has disappeared almost completely.

1.3 Gene flow and breeding activities during the formation of the Improved Awassi breed By the end of the 19th century sheep in Palestine were exclusively produced by Bedouin tribes and a few Arab villages in upper Galilee. Jewish people, starting to return to Palestine at the end of the 19th century, showed an interest in raising sheep and they bought flocks from Bedouins. In 1914 an organization called “The Shepherd” was founded to establish flocks of Awassi sheep in the Jewish settlements. In Kefar Gil’adi in upper Galilee a breeding flock was established and stud rams for this flock were purchased in Transjordan and the Jaulan (Jebel ed Druz) (Epstein, 1985). In 1932, the breeding aim to improve the Awassi sheep was formulated by the annual assembly of the Jewish sheep breeders at Kfar Gil’adi: “The improvement of milk and mutton production in sheep” (Finci, 1957). Uniform milk recording and book keeping in the flocks was adopted and in 1935 controlled mating was introduced in some flocks. In 1937 an important decision was made on improving the Awassi breed: the annual assembly at Kefar Hahoresh rejected a proposal to crossbreed Awassi with imported breeds of milk sheep in order to raise production more speedily than by selection alone (Epstein, 1985). The breeding aim was modified to concentrate on “…the increase of milk production, along with taking care to preserve the robust and healthy constitution of the Awassi breed of sheep” (Finci, 1957). In 1943, the flock book for the Improved Awassi was established and decades of in- breed selection based on this breeding goal followed. Although the Improved Awassi is a pure breed with performance gains solely from intensive and strict in-breed-selection process, during the years 1953-55 17 shipments of Hirrik sheep, totalling 14,632 ewes, arrived in Israel from Turkey, for distribution among small farmers. The animals were imported by the Israel Sheep Breeders Association and the Jewish Agency in order to overcome a shortage of Awassi sheep during that period. The Hirrik breed was chosen because of its outward similarity to the Improved Awassi. It is bred in the vicinity of Cizre on the Tigris, Kurdistan, near the borders of Syria and Iraq. In general conformation and the shape of the fat tail it resembles the Awassi, however, it is smaller than the Improved Awassi, the live weight of adult ewes varying between 40 and 45 kg. No male Hirrik sheep 318 THE IMPROVED AWASSI BREED IN ITS INDIGENOUS ENVIRONMENT were imported. The ewes were all bred to Improved Awassi rams and the traits introduced by this crossbreeding were absorbed during the following years (Epstein & Herz, 1964).

1.4 Gene flow and breeding activities to form the Assaf breed After three decades of intensive selection and breeding, the Improved Awassi had developed into a high yielding milk breed, with individual ewes producing over 1,000 kg milk per year. Nevertheless, the limited prolificacy of the Awassi breed was recognized and attempts were made to improve this trait, as well as milk production by crossbreeding with the East Friesian milk sheep. This crossbreeding program started in 1955 at Newe Ya’ar Experimental Station, near Haifa (Goot, 1986) by scientists of the Volcani Research Centre. In 1955, 23 2-year-old ewes and 12 rams, from several sires, and in 1963 another 23 ewe hoggets sired by five rams of the East Friesian milk sheep, had been imported from West Germany by ship. The subsequent crossbreeding by selection resulted in the Assaf composite breed. While pure East Friesian milk sheep were not able to cope with the Mediterranean climate and conditions, the Assaf proved to be a well adapted milk and meat dual purpose breed (Gootwine and Goot, 1996). In the following years, the high demand and prices paid for lamb meat encouraged even intensive dairy farms to produce more mutton from the dairy flocks. With its higher fertility, prolificacy, less seasonality and better growth performance (see Table 1), the Assaf suited this demand and therefore replaced more and more the Improved Awassi in intensive dairy flocks.

1.5 Gene flow and breeding activities to form the Afec Awassi and Afec Assaf lines Because lamb production contributes to more than 40% of the income of intensive dairy farms, farmers were interested in increasing prolificacy above the then level of 1.2 and 1.6 lambs born per ewe lambing for Awassi and Assaf, respectively. Crossbreeding the Improved Awassi and Assaf with Booroola Merino rams began at the Volcani Center in Israel in 1986, to introduce the B allele of the FecB (Booroola) gene, a major gene causing high prolificacy, from the Booroola Merino into the Awassi and Assaf breeds (Gootwine et al., 1992). This resulted in two new lines of Awassi and Assaf sheep, namely the Afec Awassi and the Afec Assaf. 5 Booroola Merino rams were imported from New Zealand in 1986. After 4 generations of backcrossing and intercrossing, Afec sheep carry 90% or more Awassi or Assaf genetic background and are either heterozygous (B+) or homozygous (BB) for the Booroola gene. The prolificacy of the (B+) Afec Awassi and Afec Assaf is about 1.9 and 2.3 lambs born per ewe lambing, respectively (Gootwine et al., 2000b). Today, after 15 years of breeding, a BB homozygous nucleus population has been established and BB rams are used in more than 20 intensive Awassi and Assaf flocks to produce B+ offspring which can be selected by marker assisted genotyping (Gootwine et al., 2000b).

1.6 Additional crossbreeding with the Improved Awassi and Assaf in Israel Further attempts to improve lamb production in local and Improved Awassi flocks were made during the 1972-1978 period by scientists of the Volcani Center by crossing the Awassi with the Finn and the Romanov prolific breeds (Goot et al., 1980). Although the prolificacy of the crosses was relatively high, about 1.9 lambs per lambing, the crosses were not integrated into the national flock in a sustainable way. THE IMPROVED AWASSI BREED IN ITS INDIGENOUS ENVIRONMENT 319

Over the years, several crosses between the Improved Awassi and the Assaf on one hand and rams from terminal breeds like the Australian Suffolk, American Suffolk, Dorper and Charolais on the other hand were evaluated in Israel. Results from these crossbreeding experiments did not lead to massive introduction of the new blood lines to the commercial flocks. In addition, in 1998, semen from a heterozygous CN Dorset-Hampshire crossbred ram was shipped to Israel from the University Wisconsin-Madison, USA. Assaf and Awassi ewes were inseminated and first backcross generations were evaluated at the Volcani Center experimental flock (Gootwine et al., 2002b). Although the meat leanness increased in lambs carrying the Callipyge mutation, the meat quality, especially taste and tenderness, was of lower quality and unacceptable to customers. So dissemination of the callipyge mutation was avoided.

1.7 Spread of the Improved Awassi and its crosses in Israel and the Palestinian Territories to intensive and extensive farming systems As shown in Figure 1, the results of the last 80 years of Awassi breeding in Israel are four new genotypes that emerged from the local Awassi: The Improved Awassi and Assaf breeds, and the Afec Awassi and Afec Assaf lines. The unimproved Awassi still dominates the mutton market in Israel today, with about 350,000 head raised in extensive and semi- extensive production systems, mainly by Arab sheep keepers and Bedouins. In southern Israel unimproved Awassi flocks were partly improved during the last decades by introducing breeding rams from the Improved Awassi nucleus flock of Ein Harod. The Improved Awassi with its superior milk yield dominated the intensive milk production systems until the 1980s but is nowadays replaced almost completely by the Assaf.

Unimproved 5000 BC Awassi

Hirrik ewes, 1932 AC Turkey Improved Awassi East Friesian Milk 1955 AC Sheep, Germany Improved rams Assaf

Booroola Merino, New Zealand 1986 AC

Unimproved Improved Afec Assaf Afec 2004 AC Awassi Awassi Awassi Assaf

Figure 1: Historical overview on Awassi and Assaf breeding in Israel

Today there is only one flock of the Improved Awassi – the Ein Harod flock – with about 1,500 sheep which are recorded in the national flock book. The nucleus flock of the Afec Awassi is kept in the Ein Harod flock, and the number of breeding ewes is increasing 320 THE IMPROVED AWASSI BREED IN ITS INDIGENOUS ENVIRONMENT gradually with about 300 head of Afec at the end of 2004. Recently, a project was started to introduce the Afec Awassi to the semi-extensive Bedouin flocks in the Negev. By the end of 2004 there were about 700 Afec Assaf ewes in different commercial sheep flocks and the breeding nucleus at the Volcani Center in Bet Dagan. In the Palestinian Territories the majority of the sheep stock is kept in the West Bank with a small number in the Gaza strip. The Palestinian Central Bureau of Statistics counted 590,000 Awassi sheep and 240,000 sheep of “other” breeds (mainly Assaf, some Mutton Merino and Dorper) in the Palestinian Territories in 2002/2003. In the West Bank, semi-intensive systems are concentrated in the North and tend to use Assaf or Improved Awassi purchased from Kibbutzim in Israel. According to records of the Kibbutz Ein Harod, about 400 Improved Awassi lambs (mainly ram lambs) have been transferred to the West Bank during the last 10 years for breeding. The majority of these transfers have been commercial transfers between private Palestinian sheep farmers in the West Bank and the Kibbutz, in some cases with logistical assistance from Arab merchants in Israel because of the restricted access of Palestinians to Israel and of Israelis to some areas in the West Bank. Only a few transfers have been started by Rural Development projects in the West Bank. Average prices of $400-450 per breeding sheep were paid. In 1997 another 200-300 Improved Awassi lambs were sold to a research station of the Palestinian Authority at Beit Kat, near Jenin, and recently, in September 2004, 60 Improved Awassi lambs were sold by Ein Harod to a private Palestinian farmer in the Ramallah area. The numbers of Improved Awassi lambs sold to Palestinians in the West Bank are increasing each year. The Assaf breed was introduced to the West Bank in the same way as the Improved Awassi but from many more sources in Israel, since many Assaf flocks exist in Northern Israel. The greater part of the sheep population of the West Bank is located in the South, in semi- extensive and extensive production systems with the unimproved Awassi as dominant breed. This unimproved stock is increasingly influenced by unofficial transfers of Improved Awassi sheep from the Bedouin sector in the south of Israel to the West Bank.

1.8 Description and performance of the unimproved Awassi, Improved Awassi and Assaf breeds The unimproved Awassi is a robust and vigorous, medium sized sheep of milk and mutton type. The typical phenotype is white with brown head and legs and large pendulous ears, but the colouring is very variable, so that there are also sheep with black, grey, white and spotted heads and legs (Finci, 1957). Rams are nearly always horned, ewes are commonly polled. In the course of several thousand years, the Awassi has become fully adapted to the sub- tropical environment of its extensive breeding area in the semi-arid or arid regions of southwest Asia. The traits which make the Awassi a favoured breed in these environments are its tolerance to heat and water stress as well as its resistance to enzootic pneumonia. It was also found that the Awassi is relatively resistant to copper poisoning (Epstein, 1985). It is well adapted to poor Mediterranean pasture and can compensate for under-nutrition during the dry season by using the stored energy reserves in its fat tail. The Bedouins favour its ability to lamb on the open field and its good mothering abilities. THE IMPROVED AWASSI BREED IN ITS INDIGENOUS ENVIRONMENT 321

Figure 2: Improved Awassi ewes Figure 3: Assaf stud ram

Source: Rummel et al. (2003) Source: Israel Sheep and Goats Breeders' Association (2001) Because of the strict in-breed-selection - rejecting crossbreeding with exotic milk breeds - the Improved Awassi is still a very robust breed, fully adapted to semiarid environments. The outward appearance of the improved type is similar to the unimproved, but the animals are heavier and have a much larger frame. The characteristics of the respiratory type of milk sheep are more pronounced than are the mutton features (Epstein, 1985). Performance traits of both the unimproved and Improved Awassi are given in Table 1. The relatively higher prolificacy of the Improved Awassi is mainly due to the improved feeding management in the systems those sheep are employed. According to the breeding goal formed at the beginning of the process by the Israeli sheep breeders to improve the Awassi sheep, the highest achievement of the improved genotype is the increased milk yield which on average is about 8 times higher. The decreased milk fat content in the improved genotype follows the well known negative correlation of milk yield and fat content. The Assaf breed is less described and reviewed in the literature than the Awassi. A detailed description of the development and the performance is given by Epstein (1985), Goot (1986), Gootwine and Goot (1996), Gootwine and Pollot (2002a) and Pollot and Gootwine (2004). The original design of the Assaf was a proportion of 3/8 East Frisian and 5/8 Awassi, but today the proportions of the two breeds may vary between different flocks and are generally unknown. It is a large framed breed with a white fleece of non-curly wool. The head is free of wool and usually white, but brown and black variations occur occasionally as well as spots on the extremities. Generally, the Assaf is still a heterogeneous breed. Ears are long and pendulous and ewes are hornless most of the time, while rams occasionally have horns. Assaf have a semi-fat tail with a fat base. Selection for milk yield without selection pressure on udder conformation resulted in a large, high-hanging udder. Compared to the Improved Awassi, the Assaf breed is less adapted to the harsh environment and therefore requires a higher level of management and feeding. Also its sensitivity to copper poisoning has to be considered, excluding higher proportions of poultry manure from the diet. Epstein (1985) describes a high risk of parasitic pneumonia for lambs of 3-5 months when raised with their mothers in extensive housing conditions. A summarized overview on the performance of the Assaf, is given in Table 1. 322 THE IMPROVED AWASSI BREED IN ITS INDIGENOUS ENVIRONMENT

Table 1: Performance of unimproved, Improved Awassi and Assaf sheep in Israel

Trait Performance* Unimproved Awassi Improved Awassi Assaf Height at shoulder Ewe: 68 Ewe: 73.7 - (cm) Ram: 75 (1) Ram: 85.4 (6) Mature live weight Ewe: 30-50, Ewe: 60-70 Ewe: 70 (kg) Ram: 60-90 (4) Ram: 110-120 (6) Ram: 120 (9) Prolificacy 0.6-1 (6) 1.28 (2) 1.57 (6) (lambs/ewe/lambing) Gestation (days) 155 (7) 152 (6) 146 (6) Birth weight (singles) Male: 4.63 Male: 5.4 Male: 6.21 (kg) Female: 4.23 (1) Female: 4.9 (3) Female: 5.28 (6) Daily gains of - Singles: 220-280 400 (average of male liveweight** (g/day) Twins: 190-246g (6) and female lambs) (9) Milk yield 40-60 kg/year (1, 4) 506 kg/214 days (8) 334 l in 173 days (10) Milk components (%) Fat: 7.5 Fat: 5-6 Fat: 5,5 -7 Protein: 5-5.5 (4, 5) Protein: 5-5.5 (6) Protein: 5-5.5 (9) Wool yield (kg) Ewes: 1.8 Ewes: 2.6-3.0 Ewes: 2.6 (9) Rams: 2.0-2.3 (5) Rams: 4.4 (6) * References (in brackets): 1: Hirsch (1933), 2: Epstein & Herz (1964); 3:Goot (1966); 4: Mason (1967); 5: Epstein (1977); 6: Epstein (1985); 7: Benjamin (1992); 8: Gootwine and Pollot (2000a); 9: Sheep and Goat Breeders Association (2001); 10: Pollot and Gootwine (2004) ** of lambs, average from birth to marketing in two groups.

1.9 Description and performance of the Afec Awassi and Afec Assaf lines Phenotypically, the Afec lines cannot be distinguished from the Awassi and Assaf breeds. Adult sheep are as well adapted to the conditions of the region as the original breeds. Differences in performance, as described by Gootwine et al. (2000b) and Gootwine et al. (2001) are mainly the higher prolificacy of about 1.9 and 2.3 lambs born per ewe lambing and 1.8 and 2.1 lambs born alive per ewe lambing, for heterozygous B+, Afec Awassi and Afec Assaf, respectively. Lower milk yields occurred in the first backcross generations, due to still higher proportions of Merino. In a direct comparison between the Awassi and Afec Awassi, Gootwine, et al, (2000b) report a total milk yield for the 4th backcross generation which is 94% of the Improved Awassi. In the same study, the number of lambs born per ewe lambing differed significantly (+0.66) between the Afec Awassi (B+) and the original Improved Awassi (++). Birth weights of the lambs are generally lower and the mortality to weaning 17% higher than in the Assaf and Improved Awassi (Sparim and Gootwine, 2000) due to multiple births. Triplets and quadruplets require improved feeding during pregnancy, more sophisticated and intensive supervision at birth and health management for the weaker lambs. THE IMPROVED AWASSI BREED IN ITS INDIGENOUS ENVIRONMENT 323

1.10 Conclusions Putting the history and development of the Awassi of Israel into the context of the present study, the following conclusions can be drawn: i The Improved Awassi in Israel is a successful example of within breed selection, without losing the adaptation of a breed to its natural environment. It also shows the genetic potential which can slumber in indigenous breeds. ii The development of the Assaf breed through crossbreeding with imported East Frisian Milk Sheep gives an example of a well adapted composite breed. Many crossbreeding efforts with East Frisian sheep have been conducted in Mediterranean countries aiming to improve lamb and milk production. (Boyazoglu, 1979; Mavrogenis, 1987; Salah and Galal, 1994; Gabina and Serradilla, 2000), but the Assaf is the only successful example for a new synthetic breed formed. iii The Afec Awassi and Afec Assaf lines bred through introducing the Booroola gene from New Zealand represent a very specific gene flow where only a major gene was transferred to improve a specific trait: prolificacy. This kind of single gene flow may dominate future breeding activities. iv Both the Assaf breed and the Afec lines show the impact of economic pressure on sheep breeding as well as on the gene flow in animal genetic resources. v The social structure of the agricultural sector in Israel, dominated by cooperative farm structures, had a decisive impact on the formation of the Improved Awassi and Assaf. Cooperative structures simplify breeding organisation, and the level of formal education of cooperative members was high (Liegle and Bergmann, 1994). vi The fact that the development of the Improved Awassi breed as well as its later gene flow has been initiated by the return of the Jewish people to Israel gives an example of the impact of human migration on gene flow in animal genetic resources. The genetic potential has been present in the region for thousands of years, but motivated breeders were needed to exploit it. 324 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL

2 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL The worldwide gene flow of the Improved Awassi started in the mid-1960s and continues today. After the successful development and introduction of the Assaf breed to Israel’s dairy sector in the early 1990s, Assaf genes were also exported on a large scale especially to Spain and Portugal. Figure 4 and Table 2 give an overview of the many destination countries. Until the beginning of the 1990s live animal exports dominated the transfers, but since 1992 veterinary regulations made it almost impossible to export live animals, semen or embryos from Israel to countries of the EU. Until now, the two new Afec lines of the Awassi and Assaf breeds have not been exported to other countries, but they are offered by the Volcani Center as promising breeds for foreign sheep milk and meat producers. Middle East joint regional cooperation projects include in their working proposals the evaluation of the Afec lines in several locations.

Figure 4: World wide gene flow of the Improved Awassi and Assaf breeds of sheep from Israel

United Kingdom Albania Spain Hungary Mediterranean East Europe and Countries Bulgaria Central Asia Portugal Romania Kyrgyzstan Former Yugoslavia Italy Kazakhstan Cyprus Turkey Iran

Israel Middle East

Jordan Abu Dabi Peru

Ethiopia

India New Tropical Zealand Burma Countries

Australia OkOceaniainawa

WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL 325

In the following chapters a detailed description of each transfer of Improved Awassi and Assaf breeding material from Israel is given. The information was gathered using the following methodological approach: Firstly, documents from the Ministry of Agriculture of Israel were used to identify destination countries and transfer stakeholders. Secondly, all the importing stakeholders were contacted via email and phone and informal interviews were conducted about the transfers and about the development and current state of the breed since its import. If the direct stakeholder of the transfer was not available for interviewing, key persons were identified in the country and interviewed. Where published literature was available, it was used instead of - or in addition to - the personal communication. Exports to Abu Dhabi, Burma, Bulgaria, and Iran could be identified by export protocols of the Israeli Ministry of Agriculture, but unfortunately, there was no response to requests for detailed information. Besides the documented transfers to the countries shown in Figure 4, there are several undocumented reports of extension officers who saw Improved Awassi sheep at the UNDP/FAO Dairy School of Naivasha, Kenya during the 1960s and 1970s as well as at a governmental farm in North-Eastern Somalia in 1979 (Meyn, 2005). These countries are excluded in the following overview because neither proven documents nor reliable information sources in the countries could be identified. Also secondary transfers (from 2nd to 3rd countries) are poorly documented and hard to trace so their extent remains largely unknown.

Table 2: Gene flow of Improved Awassi and Assaf breeding material from Israel

Import. Year of Breeding Price Type of Way of Exporter Actual stock Further Country Purpose Importer breed* import material US$ transfer transfer ** numbers* distribution Middle East Sharif Nazer Unknown Improve iAw 1973 400 ♀ lambs Commercial Uncle of the King of Ein Harod meat Jordan Local farmers in 100 ♀ lambs Improve milk Karak region; Ass 1998 Unknown Ass.: 200 ↑ 4 ♂ lambs & meat Jordan Ministry of crossbreeding of Jordan Grant of Truck iAw.: 563 ↑ iAw 1999 230 lambs Development Agriculture iAw rams with Israel Ein Harod Cross.: 100 ♀ lambs aid unimproved iAw 1999 Improve milk . 15 ♂ lambs Awassi production Jordan M.O. Agricult. & Ass 2004 5 ♂ lambs 1,500 Unknown DANIDA 203 ♀ lambs Pure- & Local farmers in Iran iAw 1965/66 100,100 Commercial Unknown Unknown Ein Harod Unknown 42 ♂ lambs crossbreeding Quazin area 50 ♀ lambs iAw 22,700 6 ♂ lambs Plane via Not named private Abu Dhabi 1996 Unknown Commercial Ein Harod Unknown Unknown 50 ♀ lambs Cyprus businessman from Jordan 23,300 Ass 6 ♂ lambs Gazit 50 ♀ lambs Caylanpinar Agricultural iAw 1973 24,500 Truck Local Awassi 10 ♂ lambs Improve local Commercial Institute, Sanli Urfa, Ein Harod 900,000 local Turkey herders in East Awassi Turkey Awassi → iAw 1991 7 ♂ lambs 3,150 Plane Turkey

Mediterranean Region & Western Europe Herdade do 1,000 1999 65 ♂ Unknown Improve milk Dr. Amos Bareli, Israeli Italy Ass Commercial Plane Matinho crossbreds with - -2004 10 ♂ production Veterinary, Sardinia, Italy Sarda breed ↑ 1968 810 ♀ lambs Improve milk Ministry of Agriculture, All iAw culled Cyprus iAw 365,000 Commercial Ship Ein Harod Australia -1973 21 ♂ lambs production Nicosia, Cyprus in 1993 5,000 doses M. Sarnadas, Sociedade 1991 Ass semen 50,000 Agricola da Herdade do Gazit 1993 Improve milk Several hundred All over Portugal, Portugal 260 embryos Commercial Plane Matinho, Lda. production thousands ↑ Spain, UK, Italy 1,854 doses Y. Barradas, Evora Ass 1992 18,540 Moledet semen M. Carvajal, Lisbon

Table 2 continued

Import. Year of Breeding Price Type of Way of Exporter Actual stock Further Country Purpose Importer breed* import material US$ transfer transfer ** numbers* distribution 150 ♀ 150.000- iAw 1971 82,500 Other provinces 50 ♂ Cria Ovina de Malpica 200.000 → Ship Ein Harod in Spain, 120 ♀ lambs S.A., Malpica-Tajo, Spain pure- and iAw 1973/74 106,500 Portugal 130 ♂ lambs crossbreds 1977 120 ♀ lambs Ass 59,000 20 ♂ lambs Improve milk Spain Commercial 110 ♀ lambs production Private sheep farmer in 700,000- Ass 1978 66,000 40 ♂ lambs León, Spain 800,000 ↑ Plane Gazit (purebreds and Portugal 100 ♀ lambs Ass 1979 62,000 crosses of 40 ♂ lambs various degrees) 4,000 doses Ass 1993 40,000 Mr. Carlos Guerra, Spain semen United 3,000 doses Improve milk Herdade do Ass 1997 Unknown Commercial Plane Dr. Ian McDougall, UK 100 cross.↓ - Kingdom semen production Matinho Eastern Europe & Central Asia 200 ♀ lambs Ship & iAw 1989 115,000 25 ♂ lambs land Bessenvei Cooperation, iAw 1990 200 ♀ lambs 100,000 Hajdú-Bihar County, South-East Hungary 49 ♀ lambs Improve milk 200 pure↑ Hungary iAw 1990 26,900 Commercial Ein Harod Albania 4 ♂ lambs production Plane Awassi® Rt Corporation 3500 cross.↑ Bakonszeg, Hajdú-Bihar 100 ewes iAw 1991 53,000 County, South-East hogged Hungary iAw 1998 5 ♂ lambs 3,850 Bulgaria iAw 1977-79 1,786 lambs 803,700 Unknown Commercial Ship Unknown Ein Harod Unknown Unknown 80 ♀ lambs Improve milk Ministry of Agriculture, Romania iAw 1973 51,000 Commercial Ship Ein Harod Unknown - 20 ♂ lambs production Romania Former Improve milk Ministry of Agriculture, iAw 1969/70 450 lambs 225,000 Commercial Ship Ein Harod Unknown - Yugoslavia production Yugoslavia 1,000 doses iAw 1996 10,000 Kazak Research Techn. semen Improve milk Development 30 pure↑ Kazakhstan Plane Institute of sheep Ein Harod Kyrgystan 20 ♀ lambs production aid ~200 cross.↑ 1999 10,250 breeding, Almaty region 5 ♂ lambs

Table 2 continued

Import. Year of Breeding Price Type of Way of Exporter Actual stock Further Country Purpose Importer breed* import material US$ transfer transfer ** numbers* distribution Tropical countries iAw 1980 10 ♂ lambs - Improve meat On Station 4 ♀ lambs Ministry of Agriculture, Local farmers in iAw 1984 - and coarse Development 119 pure→ Ethiopia 5 ♂ lambs Plane sheep breeding Centre, Ein Harod Highlands of wool aid 400 cross.↓ 18 ♀ lambs Debre Berhan, Ethiopia Ethiopia iAw 1994 - production 8 ♂ lambs 25 ♀ lambs iAw 1990 2,082 Nimbkar Agricultural 5 ♂ lambs Improve milk Development Very few pure India Plane Research Institute Ein Harod Regional farmers 100 doses production aid and crossbreds↓ iAw 1995 1,000 (NARI), Phaltan, India semen Development Burma iAw 1980-83 20 lambs Gift of Israel Plane Unknown Ein Harod Unknown Unknown Meat aid 11 ♀ lambs Depratment of Animal Improve milk 6 ♂ lambs Production at the 185 pure→ Peru Ass 1988 12,700 yield for Commercial Plane Gazit Regional farmers 500 doses Universidad Nacional 165 cross. ↑ lambs semen Agraria La Molina Oceania New Milk AWASSI New Zealand Offspring went to iAw 1991 143 embryos 27,800 Commercial Plane Ein Harod Only very few↓ Zealand production Ltd., Dr. Jock Allison Australia Mutton Ministry of Agriculture, ~ 100000 Australia iAw 1987 311 embryos Unknown Commercial Plane Cyprus - production Australia, Dr. Lightfoot purebreds 5,433 lambs Purebreds: 100,912 + unknown Total: 1965 1,100 doses 9 development aid numbers in diff. countries 15 iAw 2,134,032 -1999 semen 19 commercial transfers crossbreds: 204,100 + unknown countries 143 embryos numbers in diff. countries 687 lambs 1977 11,354 doses 2 development aid > 1.2 million pure- & crossbreds of 7 countries Ass 333,040 -2004 semen 8 commercial transfers various degrees 260 embryos iAw = Improved Awassi; Ass = Assaf; pure = purebred; cross = crossbred; ↑= increasing population; ↓= decreasing population; →= stable population **Ein Harod = Kibbutz Ein Harod, Awassi Fold Division, 1860 Ein Harod Ihud Gazit = Kibbutz Gazit, D.N. Emek Israel 1934; Moledet = Kibbutz Moledet, M.P. Gilboa 1913 Herdade do Matinho = Sociedade Agricola da Herdade do Matinho, Lda. P.O. Box 22. Quinta da Moutosa, 7320 Castelo de Vide, Portugal, Cyprus = Improved Awassi flock of the Ministry of Agriculture, Nicosia

WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL 329

2.1 Middle East The Middle East region is the natural environment of the Awassi breed of sheep and in most of the countries stocks of unimproved Awassi sheep can be found. The majority are kept in extensive, sometimes pastoral farming systems for mutton production. If sheep are milked it is usually for home consumption or rural markets. In several countries of the Middle East efforts and national breeding programs have been launched to improve the Awassi sheep (Jordan, Turkey). In order to supply such breeding programs with improved genetic material as well as to improve indigenous breeds through crossbreeding, the Improved Awassi of Israel has been imported to a number of countries.

2.1.1 Jordan Source: Carasso, Y., 2004; Gootwine, E., 2004 The first transfer of Improved Awassi sheep from Israel to Jordan took place in 1973, when the uncle of the Jordanian King bought 400 ewe lambs from the Kibbutz Ein Harod. The flock was transferred to his private farm in Jordan but the development and further distribution of Improved Awassi breeding material from this flock to local Jordanian farmers is unknown. In 1996, Jordan and Israel signed an agricultural cooperation agreement which apart from agricultural trade and plant protection contained a sheep husbandry project, as part of the first governmental peace project between Jordan and Israel. The aim of the project was to establish a nucleus flock of Improved Awassi and Assaf in the southern governorate of Karak, to provide Improved Awassi and Assaf breeding material to local herders. Beside the sheep flock, a dairy plant was planned to process sheep milk produced in the region. The project was implemented in 1998 by the Jordanian Ministry of Agriculture and Israel’s Centre for International Cooperation (MASHAV). In 1998, 100 Assaf ewe lambs and 4 ram lambs were transferred from Israel to Jordan, in 1999 230 Improved Awassi lambs and in a third transfer during 1999 another 100 ewes and 15 rams. All transfers were conducted by trucks. While the Improved Awassi are kept at the Karak sheep breeding station, the Assaf flock was established on the Al-Mushairfeh private farm in the same region. Both flocks are supervised by the Jordanian extension service. The costs for the sheep and the dairy plant equipment were provided by the Israeli Ministry of Agriculture, an investment totaling $500,000, while the Jordanian government provided the land, buildings for sheep shelter and the dairy plant as well as the manpower to run the farm. In its initial phase, the project was endangered because opponents of the 1994 peace treaty spread the rumor among Karak’s residents that Israel would try to buy land and dominate Jordan’s economy (Charkasi, 1999). Since 2000, the sheep breeding program (at the Karak demonstration farm) has been included in the Middle East Regional Program between Denmark, Egypt, Israel, Jordan and the Palestinian Authority, initiated by DANIDA, the Danish Agency for Development Assistance. In November 2004, another 5 Assaf ram lambs were transferred to the Assaf flock in Jordan, founded by the DANIDA project. Israel provided the animals for a price of $300 per ram, i.e. half of the commercial price. At the end of 2004, 563 Improved Awassi sheep were kept at the Karak sheep farm and about 200 Assaf sheep at the private farm Al-Mushairfeh. In 2003, 22 Improved Awassi lambs have been sold from the nucleus at Karak to local sheep farmers. The average annual milk yield in the flock of 200 milked ewes at Karak was 145, 165, 189 330 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL and 178 litres per ewe and year for the years 2000, 2001, 2002 and 2003, respectively. The main reason for the poor milk yields is the inadequate feeding regime.

2.1.2 Iran Source: Epstein, H., 1985 In 1965-66, in an effort to increase the milk production of the local Baluchi and Shal sheep in the Quazin area of Iran (situated in the northwestern corner of the central plateau, west of Tehran), Iran imported 42 male and 203 female Improved Awassi lambs from the Ein Harod flock in Israel at a price of $100,100 for pure- and crossbreeding with the two Iranian breeds. While the crossbreeding of the Improved Awassi x Shal was conducted at an experimental station with 100 sheep from each breed, the crossbreeding of the Improved Awassi x Baluchi was implemented at village level in the traditional production system as well as with an intensively managed demonstration flock.

Table 3: Performance of Improved Awassi, Baluchi and Shal breed and their crossbreds in Iran

Improved Baluchi Shal iAw x Baluchi iAw x Shal Awassi (iAw) Body weight of ewes (kg) 79 67 77 68 76 Mortality of adult ewes (%) 4.3 3.0 3.2 13.3 4.8 Average milk yield (l/year) 333 189 210 211 284 Prolificacy (lambs per ewe) 1.25 1.18 1.41 1.43 1.45 Birth weight (Singles, kg) 5 4.6 5 4.8 5.1 Mortality of lambs 0 – 180 m: 23.4 10.1 7.8 11.9 14.4 days (male &female), (%) f: 13.0 6.8 9.1 12.4 11.8 Daily weight gain lambs S: 230 S: 226 S: 231 S: 233 S: 244 (Singles, Twins, g/day) T: 236 T: 217 T: 236 T: 233 T: 236 Fleece weight of ewes (kg) 2.43 1.76 2.0 2.39 2.41 Source: Wallach & Eyal (1974), QDA (1970), both cited after Epstein (1985) Information about the further development of the Improved Awassi stock and its crosses in Iran is not available.

2.1.3 Abu Dhabi Source: Carasso, Y., 2004 In 1996 a private businessman from Jordan visited the Agritech exhibition in Israel and contacted sheep breeders in Israel. He had a request from a private sheep farmer from Abu Dhabi to import Improved Awassi and Assaf sheep. In October 1996 50 ewes and 6 rams of Improved Awassi and the same numbers of Assaf sheep were transferred by plane to Cyprus and from there to Abu Dhabi. The sheep were bought from the Kibbutzim Ein Harod and Gazit at a price of $22,700 and $23,300. The destination address in Abu Dhabi, the development of the imported stock and the current state of the flock are unknown. WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL 331

2.1.4 Turkey Source: Emsen, E., 2004 In 1973, the Improved Awassi was introduced to Turkey to raise the milk yield of local sheep through crossbreeding, and to improve the local unimproved Awassi stocks in Eastern Turkey. 50 ewes and 10 rams of Improved Awassi were bought for $400 and $450 per ewe and ram lamb respectively, by the Caylanpinar Agricultural Institute, Sanli Urfa, Turkey and transferred from Israel to Turkey. After arrival, the sheep where transferred to the Ataturk University in Eastern Turkey. In 1991, another 7 ram lambs were transferred from Israel to Turkey by plane for $3,150. After initial crosses with local breeds (Chios, Tushin, Redkaraman, Daglic, Akkaraman) and performance testing in local sheep production systems it became obvious that the Improved Awassi could not maintain its high level of milk production under rural conditions. Eastern Turkey, where the Improved Awassi was brought to, is characterized by high altitude (2000 m) and long, snowy winters. The breed had difficulty adapting to this environment, catching respiratory diseases and unable to cope with the very low level of nutrition. So in the following years, the breed was used to improve the mutton production of the local Awassi through crossbreeding. Today there are about 900,000 purebred Awassi sheep in Turkey, which have been partly improved through the national breeding program in which the imported Improved Awassi from Israel participated, representing 2-3% of the total sheep stock of the country. The number of crossbreds is unknown, but seems to be much higher. The performance of the Improved Awassi under semi intensive management in Turkey, compared to local sheep breeds, is presented in Table 4:

Table 4: Performance of Improved Awassi, Redkaraman and Tushin breed of sheep in Turkey

Improved Awassi Redkaraman Tushin Rate of ewes lambing/ewes exposed (%) 87 85 95 Litter size 1.2 1.16 1.07 Average daily milk yield (l/day) 0.57-0.78 - - Lactation length (days) 120 - - Milk components (%) Fat: 6.6, protein 5.6 - - Average daily gains of lambs (birth to 214-252 142 131 marketing) (g/day) Source: Emsen (2004)

2.2 Mediterranean Region and Western Europe The Mediterranean countries have a long tradition in breeding milk sheep and dairy processing of sheep milk. Nowhere else in the world has such a manifold diversity of local breeds of milk sheep as well as sheep milk products developed. Following the general trend in European agriculture during the last five decades the sheep milk production systems have become rapidly more intensive, using high yielding specialized exotic breeds. As part of this, the Improved Awasssi and Assaf breeds were introduced to these countries, especially to the Iberian peninsula. The Assaf succeeded dramatically, outnumbering the traditional milk breeds only a decade after its first import. Other than this and exports to the UK (see below), no transfer of genetic material of Improved Awassi breeds from Israel to Europe is known to the authors. 332 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL

2.2.1 Portugal Sources: Sarnadas,M., 2004; Roldao,D., 2004; www.herdadematinho.pt The story of the success of the Assaf in Portugal started in 1991. For several generations, Mr. Sarnadas’ family had kept local milk sheep – Serra de Estrêla – on his farm at Herdade do Matinho, in the province of Alto Alentejo near Castelo de Vide. In 1975 he imported a flock of 200 Awassi x Manchega crossbreds from Spain because he was not satisfied with the poor milk yield of the Serra de Estrela (less than 1 litre per day). In 1988 the Sociedade Agricola da Herdade de Matinho was founded to create a modern sheep dairy farm to improve productivity. New stables, milking parlours, cheese rooms and feed stores were built by Mr. Sarnadas. Several important milk breeds of Europe and the Mediterranean were investigated by the management to find the most suitable, high yielding milk breed for the new, intensive production system. The Assaf breed from Israel was finally selected, because of its high milk yield in combination with heavy, fast growing lambs and the ability to adapt to the dry and hot climate of the Mediterranean. Following this decision, 5,000 doses of semen and 260 frozen embryos were purchased and imported by the company from the Kibbutz Gazit in Israel at a value of about $50,000 between 1991 and 1993. The embryos were implanted into the Awassi x Manchega ewes and the resulting offspring inseminated with Assaf semen. So the company had a purebred Assaf flock for production and further breeding. Beside dairy production and processing, an Artificial Insemination Centre was set up by the Herdade do Matinho according to the stringent regulations of the European Union, and the company has become the main source of Assaf breeding material outside Israel. All ewes are milk recorded and genetic material (breeding rams and ewes, semen and embryos) are offered for national and international sale with detailed performance records, available on the internet (www.herdadematinho.pt). Today, 2,000 pure Assaf sheep are kept at the nucleus flock of Herdade do Matinho, with an average milk yield of 320 litres in 150 days (with single ewes up to 606 litres in 150 days) and 400-500 litres in a lactation of 200-240 days. Compared to the milk yields of the local dairy breeds of 80-120 litres and 120-150 litres per year for the Saloia and Serra de Estrêla, respectively, and 50-70 litres per year for the Merino breed, which is still being milked in some regions, these numbers show the extraordinary increase in milk yield achieved by introducing the Assaf. Parallel to the imports organized by the Herdade de Matinho in 1992 (who thought they would have the sole rights on the Assaf breed in Portugal), 1854 doses of semen were sold from another Kibbutz in Israel, Moledet, Gilboa at a price of $18,540 to two farmers in Lisbon, Mr. Y. Barradas in Evora and Mme M. Carvajal. In Portugal, Assaf rams crosses with Lacaune, Saloia, Serra de Estrêla and Merino sheep were mainly uncontrolled. On some farms, the Assaf is also produced as a purebred. Although neither numbers for purebreds nor crossbreds are available, according to local sheep breeders the Assaf became the dominating sheep breed in Portugal’s dairy sheep sector, reaching several hundred thousand sheep today (Roldao, 2004). Among the main cheese-producing areas of Portugal, in the region Castelo Branco most of the dairy sheep farmers keep Assaf in intensive production systems, mainly because of the development of the cheese industry in this region and the absence of a traditional, local milk breed. With the introduction of the Assaf breed, the extensive pastoral sheep production systems had to change dramatically into much more intensive production systems based on irrigated pasture, heavy concentrate WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL 333 supplements and a sophisticated reproduction management. Where the Assaf was kept in traditional extensive systems, it could not compete with the rustic indigenous breeds. In contrast to Spain, there is a national Assaf breeding organisation in Portugal, ACOSSAF, and a closed flock book with about 5,000 registered sheep. The organisation was created by 10 farmers immediately after the Assafs were imported in 1993, representing the only members still today. In the past ten years, thousands of doses of semen, embryos and breeding animals have been exported to Spain, with the Herdade de Matinho as the dominating importer. Assaf breeding material has also been exported to Italy (see 2.2.3) and the United Kingdom (see 2.2.5) and attempts are being made by the company to export Assaf semen and embryos to Canada. But in recent years, veterinary health export requirements became much stricter in the EU, and are difficult to fulfil in a country like Portugal where diseases like scrapie, maedi visna or Border disease are present. Because of this situation Spain recently closed its borders to sheep breeding material from Portugal, and the transfer to Canada was also not successful. Facing this problem, the Herdade do Matinho is trying to establish a closed nucleus, which is free of these diseases, especially maedi visna, in order to make further exports of breeding material possible. Beside the Improved Awassi crosses imported by Mr. Sarnadas from Spain there have been several other private imports of Improved Awassi crosses from Spain to Portugal (see 2.2.2) but compared to the success of the Assaf their numbers are negligible. Today only a few farmers keep Improved Awassi crossbreds, but tend to replace them with Assafs (Roldao, 2004).

2.2.2 Spain Source: Ugarte et al., 2001; Wellham, 1976; Epstein, 1985 In Spain, milk sheep production systems have a long tradition, producing 11% of European sheep milk today. The traditional areas of dairy sheep farming are the Autonomous Communities of Castila-León (C-L), Castilla-La Mancha and Madrid (C-M), the Basque Country (BC) and Navarre (Nav). The traditional sheep breeds in those areas are the Churra and Castellana breeds in C-L, Manchega in C-M and Laxta in BC and Nav (Gallego et al., 1994). Even though considerable efforts have been made to improve these local breeds through different breeding programs, the annual genetic gain ranged only between 1-2% per year. Therefore, more and more farmers have chosen to introduce foreign high yielding dairy sheep breeds and more intensive production systems in an attempt to increase yields in the short term. Beside the Improved Awassi and Assaf breeds from Israel, which are the dominating breeds, East Friesian and Lacaune are also present today with about 10,000 and 75,000 heads, respectively (Ugarte et al., 2001). Improved Awassi were first introduced to Spain by the Spanish sheep breeding company Cria Ovina de Malpica, S.A. (Malpica Sheep Breeding Company Ltd), located around Malpica- Tajo near Toledo, some 106 km southwest of Madrid. In 1955 the Duke of Arion took over the Malpica estates and began to improve the milk yield of the traditional Talavera breed (a Manchega x Merino cross) by crossing it with Manchega rams. However, by 1970 it became apparent that rising production costs could not be met by continuing with the existing system of sheep farming. Hence, it was decided to increase the efficiency of the milk flocks by crossbreeding with Improved Awassi rams from Israel (Welham, 1985). The Improved Awassi of Israel was chosen because, apart from high milk yields, it is a hardy animal, 334 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL adapted to the Mediterranean climate of Spain and suitable for management in large flocks. For this purpose 150 ewes and 50 rams were imported from Israel in 1971 at $82,500, and a further transfer of 120 ewe lambs and 130 ram lambs at a price of $106,500 was made in 1974/75. The 180 imported Improved Awassi rams were first mated to the purebred imported Awassi ewes and then together with a selected number of Spanish-born pure Awassi rams for a crossbreeding program. Welham (1976) estimated the number of F1 crossbreds by Awassi rams out of Talavera and Manchega ewes at 15,000 and from Churro and Castilian ewes at 25,000. But the experiment did not succeed and all the stock was sold (200 sheep to Mr. Sarnadas in Portugal). In the following years, the variety of Awassi crosses, called Malpica, spread to other provinces in C-L (especially to Zamora and Salamanca, and less to Palencia, Valadolid and León) (Ugarte et al., 2001). Wellham (1976) and Epstein (1985) gave the following performance levels of the Malpica:

Table 5: Performance of the Malpica (Awassi x Mancha, Talavera, Churro and Castilian) breed in Spain

Lactation Pure Awassi in Malpica Mancha Spain Number Milk yield 1 348 l in 232 days 180 l in 170 days 80-95 l/lactation 2 410 l in 264 days 210 l in 200 days Growth of lambs - 257 214 (g/day) Source: Epstein, 1985; Wellham, 1976 The import by the Malpica Sheep Breeding Company Ltd was the only transfer of live Awassi sheep to Spain. Later, in 1998, frozen semen entered Spain from Israel via Portugal. The current population of Awassi sheep in Spain is estimated to range from 150,000 and 200,000 sheep with a high level of crossbreeding with local breeds. The average milk yield is around 1.5 litres per day. Today, there is still no breeding program in Spain for the Improved Awassi and its crosses, the Malpica (Ugarte et al., 2001). Besides its use in the commercial dairy sector, the Improved Awassi breed was used as a reference breed in a scientific investigation which examined the genetic relationships among the indigenous Spanish breeds of sheep, Churra, Laxta, Manchega, Raza Aragonesa and Merino by genotyping for 19 DNA microsaltellites (Arranz et al., 1998). Only a few years after the Improved Awassi was imported from Israel, the Assaf was introduced to Spain in three transfers. A farmer in León imported from Israel by plane 120 female and 20 male in 1977, 110 female and 40 male in 1978, and 100 female and 40 male lambs in 1979. He bought the sheep from the Kibbutz Gazit at $400 per ewe lamb and $550 per ram lamb. Between 1985 and 1987, this first nucleus was sold to other farms in Salamanca, Zamora, León, and even to Portugal. Later, the Assaf spread from these flocks to the rest of Spain (Ugarte et al., 2001). In 1993, the Herdade de Matinho transferred 4000 doses of Assaf semen from Israel to Portugal and from there to a farmer in Spain, Carlos Guerra. This transfer was conditional to the import of Assaf embryos from Israel to Portugal by the Herdade de Matinho from the Kibbutz Gazit, because it was difficult to transfer the semen directly to Spain due to strict veterinary health import protocols. But in the following years, until Spain recently dramatically tightened up its import health requirements, this Portuguese insemination and WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL 335

Assaf breeding company sold thousands of doses of semen and about 150-200 breeding rams each year to farmers in Spain. Today the population of Assaf sheep in Spain is estimated to be around 700-800,000 animals. Not all animals are purebred, but the level of Assaf blood is probably higher than 80%. In consequence, the Assaf breed has had a significant effect on the population of Churra and Castellana sheep, and on the traditional livestock production systems. While in 1980 these local breeds numbered 1.5 Churra and 1.4 million Castellana sheep, by 1998 these figures decreased to 750,000 and 250,000 sheep, respectively (Ugarte et al., 2001). The Assaf breed is less adapted to the local environment and cannot cope as well with poor nutrition compared to the Churra and Castellana breeds. This has meant that where it is adopted, the traditional extensive production systems dependent on long periods of grazing natural pasture, fallow lands and stubble fields have had to change to an intensive – almost zero grazing – dairy system with heavy concentrate supplements and irrigated pastures close to the farm. Ugarte et al. (2001) give an average milk production of 2 litres per day and ewe for the Assaf in intensive dairy systems in Spain. In contrast to the Improved Awassi, several associations of Assaf farmers are trying to organize breeding schemes, even though there is no national breeding organization for Assaf sheep. Some are partly financed by the local government, such as the Diputación de León with 54 flocks and 25,000 ewes. There are other initiatives which are fully supported by private funds, like the ASCEGA (Association of Assaf Spanish breeders) with 27 farms and 20,000 ewes. Another one consists of 10 Camporreal farmers (Madrid) with about 11,000 ewes, and a third one consists of 12 Valladolid farmers with 10,000 animals (Ugarte et al., 2001). In Spain, there are numerous brand names of sheep cheese, the origin and quality of which is guaranteed by Denominaciones de Origen (D.O.: Label of Origin), which is linked to a specific region and local sheep breed. Nevertheless, most of the Spanish do not pay a higher price for this milk but set the price given per liter according to the contents of fat, protein, bacteria and somatic cell counts irrespective of the breed of sheep. This is one reason why farmers changed to the high yielding Assaf or Improved Awassi instead of producing less milk of higher quality with local breeds. On the other hand, it has motivated many shepherds of the Basque country and Navarra region (50%) to hand process brand name cheese on their farms from local breeds to increase their income. This is presumably the main reason for the lower level of Assaf and Improved Awassi sheep in this area (Ugarte et al., 2001). Problems of both the Improved Awassi and Assaf breeds in Spain include a higher incidence of mastitis than in local sheep breeds, and in cases of inadequate feeding, low fertility and toxemia, because of the higher energy requirements by higher milk yields. Moreover, the traditional role of the sheep as landscape conservator is lost in production systems with zero or irrigated pasture grazing. In consequence, land erosion becomes more common and natural pastures become more shrubby and wooded, drastically increasing the risk of fires during summer (Ugarte et al., 2001). 336 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL

2.2.3 Italy Source: Gootwine, E., 2004 Recently the Assaf breed of sheep also spread to Italy. In 1999, 65 Assaf rams were exported from Portugal to Sardinia, where they have been crossed by artificial insemination with the local Sarda milk breed of sheep in order to improve milk production. The transfer was started by an Israeli veterinarian, Amos Bareli, who owns an insemination centre in Sardinia. The exporting company was the Herdade do Matinho Ltd. Since 1999 there have been some more transfers of Assaf breeding rams from Portugal to Sardinia (about 10 rams). Today there are about 1,000 crossbreds between the Assaf and Sarda breeds, of various percentages of Assaf blood in Sardinia.

2.2.4 Cyprus Source: Papachristoforou, Ch., 2004, Christofides, Ch., 2004 In the 1960s, programs were launched in Cyprus to improve the milk and mutton production of the local Cyprus fat-tail breed. Beside Chios and East Friesian sheep, the Department for Animal Breeding of the Ministry of Agriculture in Cyprus imported 831 improved Awassi lambs from the Kibbutz Ein Harod, Israel, from 1968-73 at a price of $365,000. The aim was to improve the ability of Chios and its advanced crosses with the Cyprus fat-tail sheep to withstand adverse management and environmental conditions in extensive production systems (Mavrogenis, 1995). In the following years a nucleus flock was established at the Experimental Station of the Ministry of Agriculture in Nicosia; crossbreeding with the indigenous Cyprus fat tail and Chios sheep (Epstein, 1985), and performance testing started. But the Improved Awassi and its crosses could not compete with the Chios breed and its crosses in the increasingly intensive local sheep milk production systems, due to the higher prolificacy of the Chios (1.9 lambs per ewe lambing), with a similar milk yield (160-170 kg per lactation). Additional difficulties have been the seasonal breeding of the Awassi and fertility problems. In 1993 the governmental nucleus flock was culled, because there was no further intention to use the Improved Awassi and its crosses in Cyprus. Today, only very few crossbreds are producing in the country. In 1986/87, 311 embryos of Improved Awassi were exported to West Australia from the nucleus flock in Nicosia (see chapter 2.5). WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL 337

Table 6: Performance of Cyprus fat tailed and Chios sheep and their crosses with Improved Awassi in Cyprus on station

Average daily milk Milk contents Daily weight 140 day yield, lactation length, Protein (P), Fat (F) gain of lambs weight of daily milk yield (%) (kg) lambs (kg) Awassi x Cyprus 122 kg in 134 days F: 6.5 239 35.8 fat tail 0.91 kg P: 6.7 Pure Cyprus fat 65 kg in 106 days F: 6.0 185 28.5 tail 0.61 kg P: 7.1 Awassi x Chios 171 kg in 146 days F: 6.0 236 35.6 1.17 kg P: 6.1 Pure Chios 137 kg in 143 days F: 6.1 217 33.7 0.95 kg P: 6.5 References (Cyprus ARI, 1972, (Cyprus ARI, 1972, (Mavrogenis (Mavrogenis 1973, 1975) quoted 1973, 1975) quoted & Louca, & Louca, after Epstein 1985 after Epstein 1985 1979) 1979) Source: Epstein, 1985; Mavrogenis and Louca, 1979

2.2.5 United Kingdom Source: McDougall, I., 2004 Dr. Ian McDougall runs a company specializing in artificial insemination and embryo transfer services for sheep in the UK. He imported 3,000 doses of Assaf semen in 1997, which he received for a service he provided to the Herdade do Matinho in Portugal. So the Assaf was introduced to the very young sheep dairy sector of England (founded 1983). The common breed in the British dairy sheep sector is the East Friesian (or British) milk sheep and Mr. McDougall crossed the Assaf with this breed. But for several reasons (unconventional phenotype, foot problems on wet pasture, grazing difficulties in the cold and wet climate), the crossbred was not accepted by British sheep breeders. Only about 100 Assaf ewes are producing in England nowadays. Another important reason for the failure lays in the fact, that artificial insemination is not common among British dairy sheep farmers, but only semen had been imported. At about the same time, the Lacaune breed, a milk sheep from France, was introduced to Britain and crosses between this breed and the East Friesian milk sheep have been favoured by the breeders. For some years, the Assaf semen was stored by Mr. McDougall and offered free of charge to all milk sheep breeders of the British Sheep Dairy Association, but with poor feedback. In March 2004 all the semen left over was destroyed.

2.3 Eastern Europe and Central Asia In Eastern Europe there is a long tradition of producing and consuming sheep dairy products. While in the past, sheep milk production was mostly limited to extensive pastoral production systems with local breeds, after the breakdown of the socialist regimes there was some restructuring of sheep milk production, calling for more intensive dairy breeds for the newly set up intensive production units. During this transition Albania, Bulgaria, Hungary, Romania, Former Yugoslavia, Kazaksthan and Kyrgystan imported Improved Awassi breeding material from Israel. While in some countries of Eastern Europe these transfers had already begun during the socialist era, the Improved Awassi was introduced to Central Asia very recently. 338 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL

2.3.1 Hungary Source: Kukovics, S., 2004 As in other countries of Eastern Europe, milking sheep and processing sheep cheese is common among Hungarian sheep keepers. The traditional milk breeds are the Tsigai and Racka, kept in extensive production systems. Their milk yields range between 50 and 200 litres per lactation, but with high percentages of butter, fat and protein. In the former communist era, attempts had been made to establish an intensive sheep dairy sector. In Bessenyei, a village in the Hajdú-Bihar County, South Eastern part of Hungary, a cooperative farm for intensive sheep milk production was established. As a dairy sheep breed, the Improved Awassi from Israel was chosen and a long term import strategy was set up by the management of the cooperative. Between 1989 and 1991, 29 ram lambs, 449 ewe lambs and 100 pregnant ewe hoggets were imported from Israel by this cooperative. The sheep were purchased for an average price of $550-770 per breeding ram lamb, $450-500 per ewe lamb and $530 per ewe hogget. While the first transfer in 1989 went probably by ship to Koper in Slovenia and from there by road to Hungary, the three subsequent transfers were by plane directly from Israel to Hungary. But the privatisation and reorganisation in Hungary during the early 1990s caused feed shortages and management failures, and the major part of the Awassi sheep stock at the Bessenyei Cooperative farm died or was sold. In 1996, the cooperative was reorganised and transferred into the Awassi® Rt Corporation Bakonszeg. By then, the sheep stock of the farm numbered 100 ewes and 8 rams of purebred Improved Awassi and about 1200 crossbreds. Today 200 purebred Awassi and about 2500 crossbred (Awassi x Merino) are kept in an intensive, zero grazing management system and another 2500 Gyimesi Racka sheep under extensive management at the cooperative farm. Most of the milk is processed in the unit’s own dairy plant to semi-stiff kneaded cheese for export. In 1998, the Corporation imported another 5 breeding rams from Israel and further imports of Improved Awassi breeding material from Israel are planned. Beside the Awassi® Rt Corporation Bakonszeg, today there are several smaller private farms in Hungary keeping about 1000 Awassi crossbreds for milking. All of them are closely linked to the Corporation, in terms of breeding as well as in management and marketing. Compared to the national stock numbers of sheep in Hungary today (about 1.2 million head), the Improved Awassi and its crosses number still less than one percent. In the extensive and semi- extensive grazing sheep milk production systems of Hungary the Improved Awassi has not become established. Main reasons mentioned by the farmers are foot-rot problems on cold and wet pastures and difficulties in grazing natural pastures. As a socio-cultural constraint, most of the traditional dairy sheep farmers do not accept the new Awassi breed, but stick to their local sheep breeds. Nevertheless, the Sheep Breeders’ Association of Hungary has set up a special breeding plan for Awassi sheep with the primary goal to increase milk production, but the Improved Awassi or its local crossbreds are only just starting to be introduced to the broad Hungarian sheep dairy sector. Since 1970, Improved Awassi sheep have also been exported from Hungary to Albania, where they have been established in the semi-nomadic production systems. In 1996, there were about 10 flocks with about 390 Improved Awassi counted in Albania (Department of Animal Breeding and Genetics, School of Veterinary Medicine, Hannover 1998). WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL 339

2.3.2 Romania Source: Florescu, I. , 2004 In 1973 the Ministry of Agriculture in Romania bought 20 ram and 80 ewe lambs of Improved Awassi from Israel ($51,000), to crossbreed with the local Tigaie and Turcana sheep to improve milk production. In Israel, the breeders association arranged the transaction and the animals were transferred by ship to Romania where they were moved to the breeding research station in Rusetu in the county of Buzau. The crossbreds of the Improved Awassi and the Tigaie and Turcana breeds did not perform well in the mountain areas of Romania due to foot rot problems. But in regions of the plains with less then 350 mm annual rainfall per year these feet problems did not occur. Beside the crossbreeding with the two indigenous breeds, a new composite breed was created at the research station for sheep and goats in Palas, County of Constanta, by crossing the breeds Improved Awassi, Merino de Palas and East Friesian Milk sheep. Its average milk production is reported to be 200 litres per lactation and the breed is well adapted to the plains of Romania. The sheep breeding station in Rusetu had to close down in 2004 and all the Improved Awassi stock and crossbreds were sold to Mr. Ioan Sufana from Insula mare a Brailei. Beside this information, no documentation on the current state of the flocks of the Improved Awassi in Romania is available, but its influence on the sheep dairy sector of Romania seems to be negligible.

Table 7: Performance of Merino de Palas, Tigaie, Turcana and Improved Awassi in Romania

Merino de Palas Tsigaie Turcana Improved Awassi Mature body weight Ewes (kg) 60-65 43-45 38-40 58-65 Rams (kg) 100-110 70-80 60-70 90-100 Age at first lambing 20-22 24 24 22-24 (months) Lambs born per ewe 1.18-1.22 1.08-1.10 1.03-1.05 1.08 lambing Survival rate birth to 85-90 90-92 92-95 90 weaning (%) Average milk yield 0.6-0.7 0.5-0.6 0.5-0.6 1.3-2.2 (l/day) Average lactation 130-150 180-190 190-200 172-250 length (days) Milk components Fat (%) 6.5 6.4 7.0 6.6 Protein (%) 6.23 5.4 5.8 5 Solids Not Fat (%) 17-18 17-18 17-18 14 Source: Raducu, 2004, personal communication 340 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL

2.3.3 Former Yugoslavia Source: Epstein, H., 1985 In 1969 and 1970 the Ministry of Agriculture, Yugoslavia imported 450 ram and ewe lambs of Improved Awassi from Israel in order to improve the milk production of the local sheep breeds, Ovce Polje and Kosovo, by crossbreeding (Epstein, 1985). The sheep were purchased from the Kibbutz Ein Harod at a price of $225,000. No information is available on the development of the breed in the country, or its present state.

2.3.4 Kazakhstan Source: Gootwine, E., 2004 Kazakhstan is the first country of Central Asia to which the Improved Awassi breed has been exported. Before Kazakhstan was included in the UDSSR, there was a common tradition to milk sheep and produce sheep cheese, but when the UDSSR took over the sheep production sector was forced to concentrate solely on producing fine wool. Based on the tradition and market for sheep milk products in Kazakhstan, a development project was set up between Dr. Gootwine of the Volcani Centre in Israel and Prof. Kasymov and later Dr. Malmakov of the Kazak Research Technological Institute of sheep breeding in the Almaty region. The project was founded by CDR-AID, an American agency development, which supports developing countries in cooperation with scientists from Israel. The aim was to develop an dairy sheep production system based on the Improved Awassi and its crosses with local breeds. In 1996 about 1000 doses of frozen semen were transferred by plane to Kazakhstan. The semen was collected from rams belonging to the flock of the Kibbutz Ein Harod and the price of $10,000 was paid by CDR. On a former Kolkhoz in the North East of the country, in the Semipalatinsk region, local Kazak fine wool ewes were inseminated and F1 offspring were born which developed well. Unfortunately, documentation on these initial results is missing and the project was terminated. In 1998, 25 ewe lambs and 5 ram lambs were transferred by plane to Kazakhstan to establish a purebred Improved Awassi nucleus at the breeding station of the institute near Almaty. Breeding material from that nucleus served to produce Awassi crosses with Kazak Fine Wool, Kazak Fat Rump and Karakul sheep on three private farms under semi-extensive management. Today the breeding nucleus at the Kazak Research Technological Institute of sheep breeding contains about 30 purebred Improved Awassi , 6 F1 crossbred ewes (Improved Awassi x Kazak Fine Wool) and purebred local sheep. The first contemporary comparison of milk production showed the superiority of the Awassi in milk production (Table 8). WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL 341

Table 8: Milk Production of Improved Awassi, Improved Awassi x Kazak Fine Wool and pure Karakul sheep on station in Kazakhstan

Breed n Total MY in 125 Average daily Maximum days (l) MY (l) daily MY (l) Improved Awassi 15 178 1.4 4.2 F1 (iAw x KFW) 4 161 1.3 2.7 KFW 14 47 0.4 1.4 iAw= Improved Awassi, KFW= Kazak Fine Wool breed; MY= milk yield

In autumn 2003, 150 Kazak Fine Wool ewes at the Madina farm in the Konyrolen village, Panfilov district, 350 km southeast of Almaty, were inseminated with Improved Awassi semen resulting in 38 F1 ram lambs and 44 F1 ewe lambs. The same number of Kazak Fat Rump ewes were inseminated in 2003 at the Dias farm near Ulguli village, Djambulsky district, 120 km northwest of Almaty, resulting in 48 F1 ram lambs and 67 F1 ewe lambs. In South Kazakhstan, in the Aryssky district on the Ahdala farm, 160 Karakul ewes were inseminated with Improved Awassi semen in cooperation with the regional Karakul sheep breeding institute. Performance data of the different crossbreds are being collected and will be used to identify the most suitable crossbred for a semi-intensive dairy sheep production system in Kazakhstan. Until now Improved Awassi sheep introduced to Kazakhstan were limited to the scientific research station and the three farms where the on-farm research takes place. Results of performance and acceptance among local sheep breeders are yet to be seen. In November 2001, one Awassi ram and about 200 doses of frozen Improved Awassi semen was exported to Abdugani Abdurasulov, head of the reproduction department at the Kyrgyz Livestock Research Institute in Kyrgyzstan, a neighbouring country in Central Asia.

2.4 Tropical countries There has been only a limited gene flow of Improved Awassi breeding material from Israel to tropical countries. Ethiopia, India and Peru imported Improved Awassi and Assaf sheep not to improve milk production, but to increase mutton and coarse wool production of local breeds through crossbreeding. The advantage of the Improved Awassi over other exotic mutton breeds was its hardiness and ability to cope with hot climates as well as to produce at high altitudes. In the context of bilateral relationship, Improved Awassi breeding material was given to Burma by Israel.

2.4.1 Ethiopia Source: Tibbo, M., ILRI, 2004 Sheep production is a major agricultural activity with a considerable economic impact in Ethiopia. The sheep population of about 24 million is one of the largest in Africa (Woldemeskel et al., 2002) and its majority (75%) is concentrated in the northern and central highlands, raised in smallholder mixed crop-livestock production systems. Milking sheep is largely limited to very view regions (North Wello and Sekota district) (Lemma et al., 1998). The highlands are dominated by indigenous fat tailed sheep that are adapted to their specific environments, namely fat-tailed Menz sheep or the Tucur breed (Hassen et al., 2004). 342 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL

The genetic potential of the indigenous sheep in Ethiopia for meat and wool production is believed to be low, and therefore the Ethiopian Ministry of Agriculture has established a number of sheep improvement programs, all based on crossbreeding with imported breeds, assuming that in-breed selection among local breeds would be too slow. To improve the highland sheep (Menz breed), Corriedale, Hampshire and Romney rams were imported from the UK for crossbreeding (Hassen et al., 2002). But those breeds and their crosses were not accepted by local farmers, and so in 1980 the Improved Awassi breed was imported from Israel as it was assumed similar in phenotype to the local Menz sheep (fat tail). The transfer was begun by the government of Ethiopia with the support of the government of Israel as a contribution to the agricultural development of the country. Between 1980 and 1994 23 ram lambs and 22 ewe lambs have been imported in 3 transfers by plane. They originated from the Kibbutz Ein Harod in Israel and were brought to the Debre Berhan Sheep Breeding and Improving Centre, 135 km north of Addis Ababa at 2800 m above sea level. The breeding project intended to produce on station crossbred rams between the local Menz and the Improved Awassi, with 75% Awassi blood, which would be spread to the sheep producers in the highlands as terminal sires. Between 1994 and 2000, the Debre Berhan Sheep Breeding and Improvement Centre distributed about 1055 (783 for breeding and 272 for fattening) crossbred rams to farmers in most of the districts in Ethiopia, mostly in the Shewa District where the station is located. Contrary to the former crossbreeding attempt with the imported thin tailed mutton breeds, the Awassi x Menz crossbreds were well accepted by the local sheep keepers, but no records were kept and so there is no reliable information about the Awassi breed level in the smallholder mixed farming systems. Research on the performance of the Awassi breed in Ethiopia indicated that they adapted easily to the tropical highlands at altitudes between 2,700 and 2,900 m above sea level. Additionally, there was a continuous increase in mean weight at birth, weaning, and annual greasy wool weights with increasing levels of Awassi breeding (Lemma et al., 1989) on station. However, due to insufficient natural pasture and lack of supplementary feeding in the mixed crop-livestock production system where the crosses with indigenous Menz type sheep were distributed, the 75% Awassi crossbred rams were unable to adapt to such a low input system. The farmers reported that they suffered from helminth parasites. Hassen et al. (2002) found that the performance of 37.5% Awassi x Menz crossbred sheep was no better than the indigenous Menz sheep in the low input system of village conditions. Tibbo et al. (2004) reported increased adult Fasciola hepatica worm burden with increasing Awassi proportion. The performance of different genotype levels of the Awassi breed is published (Lemma et al., 1989; Lemma et al., 1991; Lemma et al., 1993; Lemma et al., 1995; Demeke et al., 1995; Demeke et al., 1998; Lemma et al., 1998; Hassen et al., 2002; Hassen et al., 2004; Tibbo et al., 2004). Their performance as affected by the level of the Awassi blood and comparison with indigenous Menz breed is given in Table 9. WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL 343

Table 9: Performance of the Improved Awassi and its crossbreds with local Menz sheep in comparison to pure Menz sheep in Ethiopia

Awassi Crossbreds Menz Proportion of Awassi (%) 100 75 50 37.5 0 Performance traits Mature body weight ewes (kg) - 41 41.4 - 31.6 Age at first lambing (months) 25.0 21.4 - - 11.6 Rate of ewes lambing/ewes exposed 74.27 71.01 62 - 97.18 Lambs born per ewe lambing 1.04 1.02 1.04 - 1.05 Birth weight (kg) 4.1 3.3 3.0 3.4 2.8 Weaning weight (kg) 22.5 18.9 16.8 12.8 11.9 Daily gains of live weight by lambs (g) 172 144 128 144 68 Average live weight of lambs at 60.0 41.0 29.6 - 24.8 marketing Greasy wool production (kg/ewe/year) - 1.33 0.98 - 0.60 Source: Tibbo, 2004 Today, there are 119 pure Awassi, 51 Awassi (75%) x Menz and 341 Awassi (50%) x Menz sheep on the three Breeding and Improvement Centres Debre Berhan, Amed Guya and Sheno. The relatively low numbers of crossbreds result from the culling and test and slaughter policy started on the stations after an outbreak of respiratory diseases in 2000 (Tibbo et al., 2001). Maedi-visna (MV), a chronic viral disease of adult sheep characterised by interstitial pneumonia, was responsible for the outbreak, to which the crossbred and indigenous Menz sheep were especially susceptible (only 48% of the imported Awassi, but 92% of the Menz sheep were infected). It appears that Menz sheep are naïve to MV infection, and it may be that a related virus was introduced to Ethiopia with imported exotic breeds from the UK or from Israel (Wolsemeskel, M. et al, 2002). In both countries MV was reported (FAO/OIE/WHO, 1991-1994). In order to control the disease, the following measures were taken: The distribution of crossbred animals to farmers for breeding was stopped and all crossbred sheep from the breeding stations had to be culled. The crossbreeding program as well as new imports of Awassi sheep into the country were stopped and the pure Awassi stock on the stations were subjected to test-and-slaughter policy for MV to establish a clean flock; pure Awassi lambs are separated immediately from their dams and fostered with cow milk. In addition, indigenous Menz sheep were purchased from farmers from “MV-free areas”. In general, the Ministry of Agriculture is also reconsidering the breeding objectives and is working on an import policy of genetic material to prevent foreign diseases being imported to Ethiopia in the future.

2.4.2 India Source: Nimbkar, C., NARI, 2004 While in Ethiopia the transfer was started by governmental institutions, in India a NGO was the driving force of the gene flow of the Improved Awassi from Israel. In 1990, the Nimbkar Agricultural Research Institute (NARI), Phaltan, district of Satara, Maharasthra State and its founder Mr. B.V. Nimbkar saw the need to improve indigenous Deccani breed of Maharashtra 344 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL and the Patanwadi breed of North Gujarat through crossbreeding with an imported high yielding milk breed. Until then research in sheep breeding in India had focused on improving the yield and quality of wool and, to a lesser degree, the yield of mutton. No attempts had been made to improve the milk yields of indigenous sheep (Nimbkar et al., 1992). Better lamb survival and growth due to higher milk yields and milk for home consumption of the shepherds were the aims behind improving the milk yield through crossbreeding. The main reason for choosing the Awassi breed was its hardiness and high milk yields, promising adaptability to Maharashtra conditions. So in May 1990, NARI imported 25 Improved Awassi ewes and 5 rams from the Kibbutz Ein Harod, Israel, at a price of $2028. They were transferred by plane to Mumbai and from there to Phaltan. In 1995, 100 straws of Improved Awassi semen were imported from Israel (Volcani Centre Bet Dagan) for $1,000, paid for by GIFRID, Germany, as a research grant. But these straws contained dead semen so were useless. In 2000 a further 100 semen pellets were collected in Syria from Syrian Awassis in a bilateral cooperation. A NARI veterinarian collected and froze the semen in Syria, and in return instructed Syrian veterinarians in the technology of freezing ram semen. Between 1990 and 2000, several crosses between the Awassi and the Deccani breed were produced and 27 pure Awassi rams, 26 Awassi (50-87,5%) x Deccani rams, 14 pure Awassi ewes and 15 Awassi (50-87.5%) x Deccani ewes were sold from NARI to farmers, research stations and companies all over India. At the Central Sheep and Wool Research Institute in Rajasthan, where some of these sheep went, several crosses with the indigenous Malpura breed were produced. But the low numbers of animals sold show the poor results of the crossbreeding program. The transfers in India did not follow any broad strategy but were isolated activities. No records about the Awassi and its crosses in local production systems are available. But on the NARI station it became obvious that although the Awassi adapted well to the hot climate, it required a high level of nutrition to reach its full potential for milk and could not manage the very poor feed resources of local sheep production systems. In addition, it was susceptible to local diseases and the conception rate at the station was very low. Therefore the crossbreeding program stopped in 2000 and new strategies for improving the Deccani sheep were considered. Today, after 15 years of breeding, Mr. B.V. Nimbkar, who initiated the transfer, considers that importing the Improved Awassi was a mistake. It had no impact on the Indian sheep production systems and did not establish at local farms. But recently the remaining stock of Improved Awassi at the NARI station is being used in a new improvement strategy for the Deccani sheep. In the year 2000, a breeding plan was started at NARI to develop a composite breed of Deccani, Garole and Bannur, three indigenous sheep breeds of India, which should carry the FecB (Booroola) gene for prolificacy. Historical accounts suggest that this major gene for prolificacy, which today is found in the Booroola Merino in Australia and the Afec Awassi and Assaf in Israel, originated in the Garole sheep of India (Turner, 1982). The Garole sheep is indigenous to Sunderban, West Bengal and is the only reported prolific sheep breed in India. Its natural habitat is swampy and very conducive to diseases, so it was expected that along with gene/genes for prolificacy, Garoles may also have genes for resistance to internal parasites, liver fluke and other diseases such as footrot (Nimbkar et al., 2002). Crossing this breed with the Deccani sheep should increase the poor prolificacy of 1.04 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL 345 of the Deccani to 1.4-1.7 lambs born per ewe lambing, without losing its adaptation to the Indian production systems. In 2001 the Improved Awassi was introduced to the breeding plan to improve the milk yield of the ewes, in order to rear properly twins and triplets, and to compensate for the reduced size due to crossing with the micro-breed Garole. At the moment the Syrian unimproved Awassi semen is introduced to the dairy type Awassi flock at NARI, because the crosses, which are going to be used in the development of the new composite breed, seem to be hardier. For the future, NARI has two parallel breeding objectives: to produce a fecund Deccani sheep, carrying the FecB gene, which can be reared in the local shepherds’ environment and to produce a fecund composite for a more intensive production system by infusion Awassi and Garole breeds into Deccani x Bannur and Bannur x Deccani ewes (Nimbkar et al., 2002).

2.4.3 Peru Source: Calle, R., 2004 The traditional breed of sheep in Peru is the Criollo, descendants from Manchega and Churra sheep transferred to Peru in the 15th century from Spain, representing 60% of the current ovine population in Peru (14 million sheep). From 1920, in the southern region of the country, the Peruvian Andean Plains, those animals were crossed with several imported improved wool breeds, especially Merino, Corriedale and Romney Marsh. From those crosses the most wide spread was the Criollo x Corriedale. But with wool prices on the world market, the breeding activities in Peru’s sheep sector concentrated on improving mutton. In order to increase prolificacy and the quantity of mutton per ewe lambing, in 1984 the Department of Animal Production at the Universidad Nacional Agraria La Molina (UNALM) imported, as recommended by the FAO in these years, prolific Barbados Blackbelly rams from Barbados (multiple births and annual polyoestrus) for crossbreeding with single birth and seasonal polyoestrus ewes (criollo crosses). But it was soon obvious that the crossbred ewes did not produce enough milk to nourish the multiple offspring properly. For this reason, the University imported 17 (11 females and 6 males) Assaf lambs and 500 doses of Assaf semen from Israel (Kibbutz Gazit) in 1988 to improve the milk yield of the crossbreds at a price of $12,700. The Assaf was preferred over other milk breeds because of its relatively high body and lamb weights. But shortly after the import the responsible scientist at UNALM, Prof. Rigoberto Calle Escobar, retired and his work was discontinued. In 1989 the offspring of the imported Assaf were sold by UNALM to private persons across Peru. Prof Calle Escobar bought several of those pure Assaf and crosses and started crossbreeding Barbados Blackbelly (pure) and Assaf in the same year on his private 1 hectare farm, called Rigoranch in Cieneguilla-Lima. The result is a new composite breed, named ASBLACK ( ¾ Assaf, ¼ Barbados Blackbelly). In the beginning, crosses between Assaf and Corriedale x Criollo sheep were also used but soon excluded from forming the new ASBLACK breed. In 2002, Prof. Calle Escobar donated all the flock (398 heads) of Rigoranch together with the infrastructure back to UNALM and it is now Prof. emeritus who manages the ASBLACK breeding flock at Rigorange de La Molina. In September 2004, the flock numbered 616 sheep, 186 pure Assaf, 165 ASBLACK and 265 pure Barbados Blackbelly sheep and is kept under intensive zero grazing management. All sheep are 346 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL individually recorded by computer and production traits are evaluated. Surplus ASBLACK rams and ewes have been sold to sheep breeders all over Peru during the last 4 years, but have not been followed up systematically. The pure-bred Assaf is not established in any production system in Peru. Assaf sheep and crossbreds sold by UNALM in 1989 were crossed without control with different breeds by local sheep breeders. Neither records, stock numbers nor production parameters of those crosses are available, since there was no feedback from the farmers where the Assaf breeding material went and the transfers had no long term breeding goal. Concerning the ASBLACK genotype, Calle (2004) reports a population of ASBLACK in the Puno Department of more than 3000 head. The ASBLACK is well accepted by local lamb producers, but no production parameters are available due to lack of monitoring.

2.5 Australia and New Zealand Source: Allison, J., 2004; Grand, P., 2004 Australia represents a country with no traditional sheep milk production systems. Sheep production focused on wool and lamb production, dominating the world market of wool production today. With no specific milk breed, Australian sheep dairy operations were based on traditional meat and carpet wool breeds (Merino x Dorset; Merino x Border Leicester), with poor milk yields of 0.75 litres per ewe and day (Anderson, 1996) or 30-120 litres per head over 90-100 day lactation periods (RIRDC 2001). In 1994, Australia imported 2,000 tonnes of sheep milk cheeses annually, valued at around $20 million (RIRDC, 2001). This growing demand and the opportunity to export sheep milk products to the world market was recognized by governmental institutions as well as private industries, and the Improved Awassi sheep was imported from Israel and Cyprus (originated also from Israel), to increase the efficiency of milk production. There have been two separate transfers of the Improved Awassi sheep to Australia, one started by governmental institutions and another driven by a private company. Sunderman and Johns (1994) gave a brief history about the governmental activities: In the early 1980s the Department of Agriculture of Western Australia began a program to import Awassi fat tail sheep in order to diversify agriculture and improve the potential for exports. Dr. John Lightfoot, now Executive Director of Animal Industries, managed this project from its conception and saw mainly three opportunities in importing the Awassi sheep: 1. A new sheep breeding industry to supply Awassi cross ram lambs, young breeding ewes and chilled carcases to the premium “fat-tail” markets in the Middle East. 2. A specialized sheep dairy industry to meet growing domestic and export demands for sheep dairy products 3. An expanded carpet wool industry with the Awassi fleece replacing carpet wool imports. In 1986 embryos were collected in Cyprus from the Ministry of Agriculture’s Awassi flock, which had originated from the two best Awassi dairy flocks in Israel, the Ein Harod and Sde Nahum. Washed and frozen embryos were preferred to live animals to prevent introducing exotic diseases to Australia. The flock in Cyprus was chosen as source for the Awassi breeding material because a very high standard of disease surveillance and reporting had been WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL 347 maintained. The frozen embryos were transferred to the Cocos Island Quarantine Station and implanted into Merino ewes. From the 311 embryos collected in Cyprus, only 51 pure Awassi lambs were born in September 1987, of which 41 survived to 100 days. Most losses are thought to have occurred during the embryo washing treatments required to remove any organisms that may have been present in the embryo flushings. The young lambs were then flown to a specially built quarantine facility in Kununurra, West Australia, and kept there until July 1990 when all the animals were transported to a larger quarantine station at Wongan Hills. The relocation was necessary for the Department to check for fibre contamination of Merino clips, which represent a serious threat to Australia’s fine wool sector. Better facilities at Wongan Hills also enabled the multiplication program to proceed. During the seven years of quarantine the sheep and their progeny were subjected to rigorous observation and testing, and prior to the release the original imports were all slaughtered and their organs examined in detail for possible diseases. At the same time, the original donor flock was re-examined to confirm its freedom from disease. Finally, in October 1990, 1640 pure Awassi and crossbred (Awassi x Merino) sheep were released from quarantine. The Awassi flock was ready for commercial development in Australia. The second transfer of Improved Awassi sheep to Australia is closely linked to the names of Dr. Jock Allison, New Zealand, and Mr. Tom Grant from Australia. At first, the import was to New Zealand, and from there the Awassi went to Australia, where almost all of their offspring are now established. Only a few Awassi sheep are still producing on farms on the northern Island of New Zealand, and their numbers are decreasing. A brief history of this transfer is given below, sourced from RIRDC (2001), and personal communication with Dr. Jock Allison: The transfer of Improved Awassi sheep from Israel to New Zealand was the idea of a scientist and sheep breeder in New Zealand, Dr. Jock Allison. At the end of the nineties he founded the AWASSI NEW ZEALAND LTD, as an instrument for importing sheep. It took him three years to develop an import protocol, and to negotiate conditions of entry of frozen Awassi sheep embryos from Israel to New Zealand. This crucial work as well as all the later effort importing the sheep to New Zealand was conducted by a New Zealand team of scientists under the supervision of Dr. Allison. In 1990, Tom Grant, an Australian farmer and business man, came into contact with Dr. Allison and took a small shareholding in the company. Soon, the Australian investments increased to 90% and the name of the company was changed to AWASSI AUSTRALIA Ltd, with Mr. Grant as chairman. After negotiations with the Israeli animal health department conducted by Dr. Allison in 1991, the company bought 65 ewes and 4 rams in Israel from the Kibbutz Ein Harod at a price of $27,800. As a requirement of the import protocol, ewes of at least six years of age were chosen, as older animals have had a longer time to manifest scrapie if it was present. The sheep were shipped to a specially built interim quarantine station (4 shipping containers, which had been modified as accommodation) in a Moshav in the Arava valley, 90 km north of Eylat in southern Israel. After two rounds of breeding conducted by the New Zealand team, 143 frozen embryos were flown to Somes Island, New Zealand, a maximum security quarantine station. There they were implanted into Romney ewes, which had been bought by the New Zealand group and after 60 days of maximum security quarantine (all the health tests passed), the pregnant recipients were transferred to a research station of the Ministry of Agriculture and Fisheries (MAF) quarantine at Bulls, near Palmerston North, where 43 purebred Improved Awassi lambs were born alive. 348 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL

But efforts to dramatically increase the numbers of purebred stock failed because artificial breeding proved difficult with the Awassi breed. After four years of quarantine in New Zealand, for commercial reasons, almost the whole stock, 64 purebred Awassi and 292 crossbreds with Romney sheep (1/2 and 3/4 Awassi), was shipped to Australia in March 1995. They were brought to a farm where they could be observed by the New South Wales (NSW) Department of Agriculture and finally transferred to Mr. Grant’s farm near Cowra were he had set up a modern sheep milking unit. The results of the first mating were poor because the Awassi crossbred rams had difficulties mounting the purebred Awassi ewes. So they had to be hand mated, which was labour- intensive. Finally Mr. Grant had over 2,000 ewes, the majority being crossbreds of various degrees. The first milkings in August 1996. By the end of the import process in 1996, 25% of the AWASSI (Aust) Pty Ltd was owned by a Saudi national, 25% by a Kuwaiti and the remaining 50% by a consortium of Australians including the Grant family. It has invested approximately $2.4 million of shareholders’ funds to date in importing the Awassi sheep and establishing facilities for sheep breeding, management and milking at its base near Cowra, in Central NSW. Part of the money came from the Rural Industries Research and Development Cooperation (RIRDC), Australia. The breeding flock in Western Australia that came from Cyprus was bought by a company named YYH Holdings, Perth, which has shares held mainly by Kuwaitis. Stock numbers were increased during the last 14 years as fast as possible in order to produce prime fat tail lamb for live export to the growing Middle Eastern market. In 2004 approximately 100,000 Improved Awassi sheep were kept by this company in Western Australia near Perth. Besides establishing the stock imported from Cyprus, YYH Holdings purchased a major share in AWASSI (Austr.) Ltd. Cowra, to access the only source of genetically different Awassi breeding material in Australia to prevent inbreeding in its Improved Awassi stock. From their side there was no interest to support the AWASSI (Austr.) Ltd. To overcome serious financial problems going along with establishing the milking unit in Cowra. As a result of this policy of YYH, difficulties in the dairy process and problems of marketing the sheep milk products forced the AWASSI (Austr.) Ltd. to shut down and to liquidate the company in 2004. In this process, 2,500 Improved Awassi sheep were sold to YYH Holdings, while only 30 remained with the Grant family in Cowra. In addition to the commercial use of the Improved Awassi, the breed was used as a reference breed in an Australian sheep gene mapping project . Generally, the Improved Awassi has had no difficulties getting established in the Australian environment. Feet problems (foot rot) are reported occasionally, when the animals are kept in a wet environment. In the very profitable Middle Eastern market of live fat-tailed lambs, YYH Holding seems to be developing into a major player in the Middle Eastern mutton market, supported by new genetic material obtained to increase its Improved Awassi stock and the solid financial support of its Kuwaiti and Saudi shareholders. WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL 349

2.6 Conclusions Looking at the gene flow of the Improved Awassi and Assaf breeds of sheep from Israel during the last 50 years and the development of the breeds in the different countries of destination, the following conclusions can be drawn in the context of the present study: i In general, the gene flow of the Improved Awassi and Assaf breed of sheep is characterized by free animal movements, based on commercial interests, with a minimum of governmental impact. ii With a few exceptions, where the transfers were part of a bilateral cooperation program, the movements of the Improved Awassi and Assaf sheep can be characterized as commercial transfers with a freely agreed benefit sharing by both stakeholders, exporters and importers alike. In all cases animals were purchased. Although Israel has an interest in exporting developed breeds, most of the transfers were started by importers aiming to improve their national production systems. Development projects, emergency aid and government institutions played a negligible role in the gene flow of the Improved Awassi. iii The most successful developments of transferred Improved Awassi and Assaf breeds were organised by private persons, breeders or companies (Portugal, Spain, Australia, Hungary), while those organised by governmental institutions and NGOs seem to have had less sustained impact on the national sheep sectors (Ethiopia, India, Jordan, Iran). iv Concerning the geographical direction of the transfers, this case study on the improved Awassi breeds of sheep describes gene flows from South to North, North to South and from South to South. In terms of numbers of animals established the emphasis was on the South to North movement. Today the majority of Improved Awassi and Assaf stocks are kept in European countries, north of Israel. Farmers of the Iberian Peninsular, where the Assaf is the dominating milk sheep, had the highest benefit from the past gene flow of Awassi and Assaf breeding material. While the South to North movement is dominated by the Assaf breed, in the movement to the South the Improved Awassi plays the major role (Australia). v In the case of the Mediterranean countries, Improved Awassi and Assaf imports had a serious impact on the traditional breed structure of milk sheep. With increasing numbers of Assaf and Awassi during the past decade, the numbers of traditional milk breeds like the Serra de Estrela in Portugal and the Churra and Castellana sheep in Spain has decreased dramatically. But since those indigenous breeds are still not endangered, the Awassi and Assaf can be considered as contributions to the genetic diversity of these countries. In other places, like Australia, the transfer had no impact on the mutton breed structure at all, but the new breed exploited a specialized market (Middle Eastern premium fat tail lamb market), to which no traditional breed fitted. The latter example shows the feedback gene flows can have on the regions from which they originate: the established Improved Awassi stock of Australia already supplies the Middle Eastern market with premium fat tail lamb through considerable live exports to these countries, which will increase with growing stock numbers in Australia, and this competes with local sheep breeders in the Middle East. 350 WORLDWIDE GENE FLOW OF THE IMPROVED AWASSI SHEEP FROM ISRAEL

vi An important stimulus for the gene flow of the Improved Awassi and Assaf breeds of sheep was the development and spread of Artificial Insemination (AI) and Embryo Transfer technology, creating an efficient tool to replace expensive live animal transfers and to reduce the threat of importing foreign diseases. The rapid increase of the breed on the Iberian Peninsula would not have been possible without those technologies. Similarly, introducing the Improved Awassi via live animal import to Australia would probably have been denied by the national veterinary departments, in order to prevent foreign diseases being imported with them and threatening the national sheep production sector. vii An additional aspect of gene flows in animal genetic resources can be seen in the transfers of Improved Awassi and Assaf breeding material to Jordan and to the Palestinian Territories. In addition to their economic benefits for the local sheep breeders, those transfers have been part of an ongoing peace process and international cooperation. Sharing achievements, in this case in animal production, can build confidence. But as seen in Jordan, politically motivated actions are often less successful than commercial ones. viii With intensive management on station, the Improved Awassi showed a higher level of performance in most countries in milk and mutton compared to indigenous breeds. But in several examples, especially in the tropical countries India and Ethiopia, but also in Eastern European countries, the Improved Awassi could not maintain its level of performance in local production systems with low input. This shows the importance of production system analysis before introducing a foreign breed or crossbred, in order to prevent a mismatch of breed and production system. ix Gene flow of animal genetic resources often includes a flow of animal diseases. This can be seen on the example of Ethiopia. Local breeds are often not adapted to exotic diseases. In developing countries, where the need to raise the performance of local breeds in order to provide food is high, import regulations and restrictions have often been insufficient. x On the other hand, government restrictions and import policies can be major obstacles to gene flow, as seen in Portugal and Australia. In Portugal, exporting Assaf breeding material is increasingly difficult due to tightened European laws for genetic transfers. In the case of Australia, governmental restrictions and veterinary requirements increased the organisational and monetary investments dramatically, restricting animal transfers only to people or organisations with sufficiently high financial capital. xi Examples for a lack of fit to the production environment are the attempts of introducing the Improved Awassi to India and Romania, while the cases Ethiopia and Britain illustrate a lack of fit between breed and production system or environmental/climatic conditions. xii It is interesting that the Improved Awassi and Assaf gene flow is bilateral: while Awassi and Assaf genes contribute to the gene pool in many countries, Awassi and Assaf absorb new traits controlled by new alleles originated from other breeds, such as the Booroola gene. REFERENCES 351

3 REFERENCES Anderson, J. 1996. Opening of Awassi Australia’s sheep dairy. Media Release DPIE 96/101A, 11 October 1996. Minister for Primary Industries and Energy. Ayelet, G., Roger, F., Tibbo, M. and Tembely, S. 2001. Survey of maedi-visna (MV) in Ethiopian highland sheep. The Veterinary Journal 161: 208–210. Arranz, J.J., Bayón, Y, Primitivo, F.S. 1998: Genetic relationships among Spanish sheep using microsatellites. Animal Genetics, December 1998, vol. 29, no. 6, 435-440 (6), Blackwell Publ. Becker, D. 1958. Sheep breeding. The Sheep Breeders' Association of Israel Merhavya (in Hebrew). Quoted after Epstein, 1985. Benjamin, R.W. 1992: Sheep husbandry for lamb production in a semi-arid mediterranean environment. In: Alberda, Th., van Keulen, H., Seligman, N.G., de Wit, J. (Eds.), Food from dry lands. Kluwer Academic Publisher, 83-97. Boyazoglu, J.G., Casu, S., Flamant, J.C. 1979. Crossbreeding the Sardinian and East Frisian breeds in Sardinia. Ann. Génét. Sél. Anim. 11(1): 23-51. Quoted after Gabina and Serradilla, 2000. Charkasi, D. 1999. First governmental peace project realised. Jordan Times, 22.11.1999, Amman. Department of Animal Breeding and Genetics, School of Veterinary Medicine, Hannover 1998. http://www.tiho-hannover.de/einricht/zucht/eaap/descript/697.htm Demeke, S., Kebede, B. and Lemma, S. 1998. Carcass characteristics of Menz sheep and their crosses with Awassi and Corriedale breeds reared under different feeding regimen. Ethiopian Society of Animal Production, Addis Ababa. Women and animal production. Proceedings of the sixth annual conference of Ethiopian Society of Animal Production. ESAP Proceedings, Addis Ababa, Ethiopia. 208-216. Demeke, S., Thwaites, C.J. and Lemma, S. 1995. Effects of ewe genotype and supplementary feeding on lambing performance of Ethiopian highland sheep. Small Ruminant Research 15(2): 149-153. Epstein, H. 1971. The origin of domestic animals of Africa. African Publ. Corp. New York. Epstein, H. 1977. The Awassi Sheep in Israel. World review on animal production 13 (2). Epstein, H. 1982. Awassi sheep. World Animal Review 44: 9-18. Epstein, H. 1985. The Awassi sheep with special reference to the improved dairy type. FAO, Rome. Epstein, H., Herz, A. 1964. Fertility and birth weights of sheep in a subtropical environment. Z. Tierzücht. Züchtungsbiol. 80: 232-266. Fái, I. 1979. From the history of the sheep branch on the Jewish farms. Hanoked 41(86):11-27 (in Hebrew). Quoted after Epstein, 1985. Finci, M. 1957. The Improvement of the Awassi breed of sheep in Israel. The Weizmann Science Press of Israel, Jerusalem. 352 REFERENCES

Gabina, D., Serradilla, J.M. 2000. Sheep and goat genetic improvement strategies. North and East Mediterranean. In: Guessous, F. Rihani, N., Ilham, A. (Ed.), Livestock production and climatic uncertainty in the Mediterranean. Proceedings of the joint ANPA-EAAP- CIHEAM-FAO symposium, Agadir, Morocco, 1998. Wageningen Pers., Wageningen, 193-202. Gallego, L., Torres, A., Caja, G., 1994. Ganado Ovino: Raza Manchega. Mundi-Prensa, Madrid. Quoted after Ugarte et al., 2001. Goot, H. 1986. Development of Assaf, a synthetic breed of dairy sheep in Israel. Proceeding of the 37th annual meeting of the European Association for Animal Production, Budapest, 1-29. Goot, H. 1966. Studies on the native Awassi sheep and its crosses with the exotic East Frisian milk sheep. Rehovot, Israel. The National University of Agriculture, Volcani Institute of Agricultural Research. Pamphlet No. 115 (in Hebrew). Quoted after Epstein, H. (1985). Goot, H., Foote, W.C., Eyal, E, Folman, Y. 1980. Crossbreeding to increase meat production of the native Awassi sheep. Volcani Center Special publication No. 175.). Gootwine, E., Goot, H. 1996. Lamb and milk production of Awassi and East-Friesian sheep and their crosses under Mediterranean environment. Small Ruminant Research 20, 255- 260. Gootwine, E., Pollot, G.E. 2000a. Factors affecting milk production in Improved Awassi dairy ewes. Animal Science 71, 607-615. Gootwine, E., Pollot, G.E., 2002a: Factors affecting the milk production of Assaf dairy sheep in Israel. Proceedings of the 7th World Congress on Genetics Applied to Livestock Production, August 19-23, 2002, Montpellier, France. Gootwine, E., Braw, R., Bor, A. 1992. Lamb and milk production in Awassi, Assaf, Booroola-Awassi and Booroola-Assaf sheep in Israel. Proceeding of the New Zeeland Society of Animal Production 52, 203-205. Gootwine, E., Rosov, A., Bor, A., Yossafi, S., Zenue, A. 2002b. Introgressing of the callipyge mutation into the Assaf fat tail breed. CIHEAM: Otions mediterranéennes Vol 4. Gootwine, E., Zenu, A., Bor, A., Yossafi, S., Rosov, A., Pollot, G.E. 2000b. Lamb and milk production in the Afec Awassi and Afec-Assaf strains carrying the B allele of the FecB fecundity gene. 51th Annual Meeting of the European Association for Animal Production (EAAP), The Hague, The Netherlands. Gootwine, E., Zenou, A., Bor, A., Yossefi, S., Rosov, A. and Pollot, G.E. 2001. Genetic and economic analysis of introgression the B allele of the FecB(Booroola) gene into the Awassi and Assaf dairy breeds. Livestock Production Science 71: 49-58. Hassen, Y., Sölkner, J. and Fuerst-Waltl, B. 2004. Body weight of Awassi and indigenous Ethiopian sheep and their crosses. Small Ruminant Research 55 (1-3): 51-56. Hassen, Y., Sölkner, J., Gizaw, S. and Baumung, R. 2002. Performance of crossbred and indigenous sheep under village conditions in the cool highlands of central-northern Ethiopia: growth, birth and body weights. Small Ruminant Research 43(3): 195-202. Hirsch, S. 1933. Sheep and goats in Palestine. Palestine Economic Society, Tel-Aviv. REFERENCES 353

Lemma, S. 1995. Comparison of oats/vetch fodder crop and natural pasture for fattening Ethiopian highland sheep. Ethiopian Society of Animal Production, Addis Ababa. Proceedings of the third national conference of the Ethiopian Society of Animal Production. ESAP Proceedings. Addis Ababa: Ethiopian Society of Animal Production, 240-244. Lemma, S., Awgichew, K., Tiyo, D. and Fletcher, I. 1993. Body weight and wool production responses to supplementary feeding in Menz sheep and their crosses with Awassi and Corriedale. Institute of Agricultural Research, Addis Ababa (Ethiopia). Proceedings of the fourth national livestock improvement conference. IAR/NLIC Proceedings (Ethiopia). No. 4, 119-126. Lemma, S., Awgichew, K., Worku, G., Kitila, A. and Fletcher, I. 1989. Comparative evaluation of Menz and Awassi x Menz crossbred sheep: 1. Birth weight, weaning weight, and wool production. Institute of Agricultural Research, Addis Ababa (Ethiopia). IAR proceedings. Second national livestock improvement conference, 82- 86. Lemma, S., Awgichew, K., Worku, G., Kitila, A. and Fletcher, S. 1991. Comparative evaluation of Menz and Awassi X Menz crossbred sheep: 2. Reproduction and mature body weight. Institute of Agricultural Research, Addis Ababa (Ethiopia). IAR proceedings. Third national livestock improvement conference. IAR/NLIC Proceedings (Ethiopia). MIC – 78025. No. 3, 78-81. Lemma, S., Gizaw, S., Deresa, A., Hasen, Y. 1998. Sheep and goat research in the Amhara region. Seboka, B. (ed.); Deresa, A. (ed.). Ethiopian Agricultural Research Organization, Addis Ababa. Agricultural research and technology transfer attempts and achievements in northern Ethiopia.. Addis Ababa (Ethiopia): EARO, 202-233. Liegle, L., Bergmann, T. 1994. Krise und Zukunft des Kibbutz. Vom Wandel einer genossenschaftlichen Wirtschafts- und Lebensform. Juventa Verlag, München. Mason, I.L. 1967. The sheep breeds of the Mediterranean by C.A.B. in arrangement with the FAO. Mavrogenis, A.P. 1987. Genetic improvement of sheep in Cyprus by selection and/or crossbreeding. Proceeding of Increasing Small Ruminant Productivity in Semi-arid areas, E.F. Thomson and F.S. Thomson (Ed.), ICARDA, Aleppo, Syria, Kluwer Academic Publ. Dordrecht. Mavrogenis, A.P. 1995. Breeding systems and selection strategies for sheep improvement in Cyprus. Options mediterraneénnes, Vol.2, CIHEAM, Paris. Mavrogenis, A.P., Louca, A. 1979. A note on some factors influencing post-weaning performance of purebred and crossbred lambs. Animal Production 29: 415-418. Meyn, K., 2005. Personal communication. Nimbkar, C., Ghalsasi, P.M. 1992. Observations on the Performance of the first flock of Improved Awassi sheep in India. In: Recent Advances in Animal Production, Proceedings of the Sixth AAAP Animal Science Congress, Vol. 3, Bangkok. Nimbkar, C., Ghalsasi, P.M., Walkden-Brown, S.W., Kahn, L.P. 2002. Breeding Program for the genetic improvement of deccani sheep of Maharrahtra, India, Proceeding of the 7th World Congress on Genetics Applied to Livestock Production, August 19-23, 2002, Montpellier, France. Pollot, G.E., Gootwine, E. 2001. A genetic analysis of complete lactation milk production in Improved Awassi sheep. Livestock Production Science 71: 31-47. 354 REFERENCES

Pollot, G.E., Gootwine, E. 2000. The genetics of dairy production from the complete lactations of Improved Awassi sheep. Proceeding of the EAAP conference, The Hague, 1-10. Pollot, G.E., Gootwine, E., 2004: Reproductive Performance and Milk Production of Assaf Sheep in an Intensive management System, Journal of Dairy Science 87: JDS 4177 Take H1215. QDA (Qazvin Development Authority). Development of sheep farming in the Quazvin area. Qazvin, Iran. (Mimeo, quoted after Epstein, 1985). Raducu 2004. personal communication. General Director of the Institute for Research and Development for sheep and goats in Palas – Constanta, Romania. [email protected]. RIRDC 2001. Tom Grant’s Awassi Sheep Milk, Rural Industries Research and Development Corporation, Barton, Australia, www.rirdc.gov.au. Rummel, T., Gootwine, E., Valle Zárate, A. 2003. Analysis of sheep farming systems in Israel. Cuvillier Verlag, Göttingen. Salah, E., Galal, E. 1994. Strategies for the genetic improvement of fat-tail sheep in the Near East. In: Galal, E.S.E. and Gürsay, O. (Ed.): Strategies for the development of fat-tail sheep in the Near East.EAAP Publication No.68, Wageningen. Sheep and goats breeders association 2001. Assaf, the modern dairy sheep. Pamphlet, Tel Aviv, Israel. Sparim, I., Gootwine, E. 2000. Economic evaluation of breeding for higher prolificacy in Awassi flocks. Internal paper, Bet Dagan, Israel. Sunderman, F., Johns, M. 1994. Awassi fat tails- a chance for premium exports. Journal of Agriculture, Western Australia 35 (3): 99-105. Tibbo et al. Genetic parameter estimates for pre- and post-weaning growth performance and survival rates in indigenous Ethiopian Horro and Menz sheep breeds. Unpublished. Tibbo, M., Aragaw, K. and Deressa, A. 2004. Effects of anthelmintics and supplementation on productivity of Menz and Menz-Awassi crossbred sheep with sub-clinical helminthosis. Ethiopian Veterinary Journal 8(2): 1-22. Tibbo, M., Woldemeskel, M. and Gopilo, A. 2001. An outbreak of respiratory disease complex in sheep in central Ethiopia. Tropical Animal Health and Production, 33(5): 355-365. Turner, H.N. 1982. In “The Booroola Merino”, Proceedings of the Workshop, CISRO, Australia, 1-7. Ugarte, E., Ruiz, R., Gabina, D., Beltrán de Heredia, I. 2001. Impact of high yielding foreign breeds on the Spanish dairy sheep industry. Livestock Production Science 71: 3-10. Wallach, E., Eyal, E. 1974. The performance of intensively managed indigenous Iranian sheep and of Awassi sheep imported to Iran from Israel., Zeitschrift für Tierzüchtung und Züchtungssbiologie 91(3): 232-239; (4): 317-326. Welham, M. 1976. Crossbreeding sheep for milk and meat in a Mediterranean environment. World Animal Review 19: 24-27. Welham, M. 1985. Crossbreeding sheep for milk and meat in a Mediterranean environment. FAO Animal Production and Health Paper 55: Small ruminants in the Near East, Vol.2, 1985, Rome. REFERENCES 355

Woldemeskel, M., Tibbo, M. and Potgieter, L.N.D. 2002. Ovine progressive pneumonia (maedi-visna): an emerging respiratory disease of sheep in Ethiopia. Deutsche Tierärztliche Wochenschrift 109: 486–488.

356 CONTACT ADDRESSES

4 CONTACT ADDRESSES

Rummel, Tobias, Dipl.-Ing.agr. Anne Valle Zárate, Prof. Dr. Dorfstr. 38, 09326 Geringswalde, Institute of Animal Production in the Tropics OT Neuwallwitz, Germany and Subtropics, University of Hohenheim, Phone: ++49 3738271722 70593 Stuttgart, Germany Phone: ++497114594210, Fax: ++497114593290 Email: [email protected]

Gootwine, Elisha, Dr. Institute of Animal Science, A.R.O., The Volcani Center, P.O. Box 6, Bet Dagan 50250 Israel Phone: ++97239683752, Fax: ++97239603678 Email: [email protected]

Allison, Jock, Dr. Awassi Rt. Abacus Biotech, 4164 BAKONSZEG, PO Box 5585, Hunyadi út 83., Hungary Dunedin, New Zealand Phone: (54) 513-000/001/002; Phone: ++64 3 4776375, Fax: (54) 513-003 Fax: ++64 3 4776376, www.awassi.hu Mobile 64 21 363337, Email: [email protected] Email: [email protected]

Calle Escobar, Rigoberto, Prof. Emeritus Carrasso, Yosef Sheep and Goat Division, Extension Service Universidad Nacional Agraria La Molina, Ministry of Agriculture and Rural Telf. 3495747 anexo 164. Development, P.O. Box 28 Francisco Del Castillo 573 San Antonio, Bet Dagan 50250, Israel Miraflores, Lima –Perú Phone: 97239485306, Fax: 97239485614 Phone: 4473115 Mobile: 97252993656 Fax: 2424364 Email: [email protected] Mobile: 99857639 Email: [email protected]

Emsen, Ebru Florescu, Irina Department of Animal Science, Counsellor, Division for European Ataturk University, Turkey Integration Email: [email protected] Ministry of Agriculture, Romania Email: [email protected] CONTACT ADDRESSES 357

Grand, Phillip Kukovics, Sandor, Dr. Cowra Cheese, Research Institute for Animal Prod. and 1078 Mid Western Highway, Nutrition Cowra, NSW 2794 Australia Geszteneyés u.1. Herceghalom, Phone: ++61 0417479716 H-2053 Hungary Phone: ++3623319133 email: [email protected]

Malmakov, Nurlan, Dr. McDougall, Ian Research Institute of Sheep Breeding 48 Bath Road, Research Centre for Animal Production and Stroud GL5 3JL, UK Veterinary Phone:++447855262308 Mynbaevo village, Djambul District Email: [email protected] Almaty region, 483174, Kazakhstan Phone: ++73272214235 Email: [email protected]

Nimbkar, Chanda Papachristoforou, Ch. Nimbkar Agricultural Research Institut, Christofides, C. Animal Husbandry Division, Ministry of Agriculture, 1412 Nicosia, Cyprus P.O. Box 23, Phaltan 415523 Phone: 35722408639, Fax: 35722408656 Maharashtra, India Email: [email protected] Email: [email protected] [email protected]

Roldao, Daniel Tibbo, Makros Sociedade Agricola da Herdade do Matinho, Lda. International Livestock Research Institute ILRI P.O. Box 22. Quinta da Moutosa, Animal Genetic Resources 7320 Castelo de Vide, Portugal PO Box 5689 Phone: ++351 245901145/ 245993225 / Addis Ababa, Ethiopia 245202202 Email: [email protected] Fax: ++351 245901145 www.herdadematinho.pt Email.: [email protected], [email protected]

358

359

GENE FLOW IN ANIMAL GENETIC RESOURCES. A STUDY ON STATUS, IMPACT AND TRENDS

Editors: Anne Valle Zárate, Katinka Musavaya and Cornelia Schäfer

Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, Germany

ANNEX 9.7

361

UNIVERSITY OF HOHENHEIM Institute of Animal Production in the Tropics and Subtropics

History and worldwide development of Anglo Nubian goats and their impacts in smallholder farms in Bolivia

A. Stemmer, C. Gall, A. Valle Zárate

TABLE OF CONTENTS 363

TABLE OF CONTENTS

List of tables 364

List of figures 364

Abbreviations 364

Executive summary 365

1 Introduction 367

2 Development and distribution of the Anglo Nubian breed 368 2.1 Description of the Anglo Nubian breed 368 2.2 Origin and breed development of the Anglo Nubian 368 2.3 Distribution of the Anglo Nubian 370 2.4 Conclusions 373

3 The keeping of Anglo Nubian goats in Bolivia: History, development and impacts 375 3.1 Conditions of goat keeping in Bolivia 375 3.2 Introduction of Anglo Nubian goats to Bolivia 375 3.3 Spread of Anglo Nubians in Bolivia 376 3.4 Development of Anglo Nubian populations in Bolivia 378 3.4.1 Cochabamba 378 3.4.2 Chaco 379 3.4.3 Tarija 380 3.4.4 Chuquisaca 380 3.4.5 Potosi 380 3.4.6 Santa Cruz 381 3.4.7 Comparison with other exotic goat breeds 383 3.5 Impacts of the keeping of Anglo Nubians 383 3.5.1 Impact on goat holders 383 3.5.2 Impact on the Criollo goat population 384 3.5.3 Impact on the environment 385 3.6 Conclusions 386

4 References 388

5 Contact addresses 393 364 LIST OF TABLES

LIST OF TABLES Table 1: Introduction of Anglo Nubian goats to Bolivia and spread within the country 377 Table 2: Development of Anglo Nubian populations in Bolivia 382

LIST OF FIGURES Figure 1: Gene flow of the Anglo Nubian goat 373

ABBREVIATIONS ACCT Asociación de Criadores de Cabras, Tarija (Association of Goat Breeders, Tarija) CEDEAGRO Centro de Desarrollo Agropecuario (Center for Agronomical Development), Bolivia DRIPAD Desarrollo Rural Integrado y Participativo en Areas Deprimidas (Rural Integrated and Participatory Development in Depressed Areas) g gram IBTA Instituto Boliviano de Tecnologia Agropecuaria (Bolivian Institute of Agronomical Technology) m Meter NGO Non-Governmental Organisation PDAR Programa de Desarrollo Alternativo Regional (Regional Programme for Alternative Development) PMA Programa Mundial de Alimentos (World Food Programe) PRODIZAVAT Programa de Desarrollo Integral de la Zona Andina y Valles Altos de Tarija (Integrated Development Programme for the Andean Zone and High Valleys of Tarija) SONU Sociedad Nueva WFP World Food Programme EXECUTIVE SUMMARY 365

EXECUTIVE SUMMARY The Anglo Nubian is an example of a breed developed by combining genetic resources from different parts of the world joining performance and adaptation to tropical conditions. It is a dual-purpose goat used for milk and meat production. Its origin can be traced to Nubian goats and the Indian dairy breed Jamnapari, which may have a common ancestor in ancient Iran. The Nubian group includes the Zaraibi, Damascus and Sudanese Nubian. Herdbooks of the Anglo Nubian are kept in Britain, the USA, Canada and Australia. The countries of the South with purebred or crossbred Anglo Nubians have no such official records and information on numbers is scarce or non-existent. The initial crossbreeding, which led to the formation of the Anglo Nubian breed took place during the latter half of the 19th century with more directed efforts from 1896 up to 1910. Afterwards, the Anglo Nubian was developed by in-breed selection alone. In 1910, the Anglo Nubian was recognized as a breed in England and registry began. Anglo Nubians were exported for the first time from Britain to the USA in 1909, reaching a total of about 30 goats up to 1950. Here, Anglo Nubians were bred and selected without any further crossbreeding with other breeds. In Canada, a breeding programme was established in 1921, based on imports of Anglo Nubians from Britain. Offspring were imported into the USA and continued to have a great impact on Anglo Nubians there until the late 1940’s. From the USA Anglo Nubians were exported to Puerto Rico and Latin America as early as the 1940’s. Later on, Anglo Nubians were exported from Britain and the USA in several development efforts in Latin America, Africa and Asia. In some countries, Anglo Nubians continue to be kept as purebreds (Mexico, Brazil, Peru, Colombia, several Caribbean states, Egypt, Israel, Oman, India, Bangladesh, The Philippines, Mauritius and Malaysia) although numbers are sometimes so small that it is difficult to preserve the population (Venezuela, Ecuador, Thailand). More widespread is the use of the Anglo Nubian in crossbreeding. In some countries, like Cuba, imported Anglo Nubians together with other specialized breeds caused a decline in number of the local Criollo goat. In Bolivia, goats are mainly kept in the inter-Andean valleys at altitudes ranging from 1,000 to 3,000 m altitude. Anglo Nubian goats were imported to Bolivia in the late 1960’s up to the present. Animals originated in Argentina, Brazil, Paraguay or the USA; semen was introduced from Germany. There have also been exchanges between different parts of the country. Anglo Nubians were introduced in the regions with a tradition in goat keeping, namely the departments of Cochabamba, Tarija, Chuquisaca, Potosi, and some parts of Santa Cruz. The majority of importations and exchanges within Bolivia were planned and handled by development agencies (6 cases), non-governmental organizations (3 cases) and universities (2 cases) while only in few cases, individual goat owners imported Anglo Nubians. Introduction was successful in intensive or semi-intensive production systems, whereas no benefits were observed in semi-extensive and extensive production systems. Here, in cases of persistence of the breed, it could often not make use of its production potential and was therefore not superior to Criollo goats. However, introducing Anglo Nubians in intensive production systems had positive side effects through improved pasture and herd management. The impact of the introduction and 366 EXECUTIVE SUMMARY promotion of the Anglo Nubian in Bolivia on the local goat population in predominantly very extensive production conditions was largely unsuccessful and unsustainable. The focus on Anglo Nubians in development projects and research resulted in a waste of resources invested and limited the allocation of resources for Criollo goat improvement but did not cause direct damage. Only in exceptional cases, under more intensive management conditions, positive economic and environmental effects could be stated.

INTRODUCTION 367

1 INTRODUCTION The present case study outlines development of the Anglo Nubian goat in Britain and follows up the original transfer of the founder breeds to Britain in the 19th century. The Anglo Nubian is found in many parts of the world toay. An overview of the worldwide spread of the Anglo Nubian from Britain to USA and Canada, later to Africa and Asia as well as Latin America is given. It is described how the transfers were made and the key players are identified. The impact of the Anglo Nubian in Bolivia is examined as an example of the introduction of an improved, high performance tropical goat, to predominantly very extensive smallholder production conditions The flow of breeding animals into the country and their spread within Bolivia is described. The main actors and driving forces for the introduction and spread are identified and the impact of success and failure on the local population of goat keepers, the environment and biodiversity of goats are shown. The factors that played a role in affecting positive and negative impacts are identified where possible. Information is compiled through available project reports, literature, statistical records where available and accessible and interviews with experts.

368 DEVELOPMENT AND DISTRIBUTION OF THE ANGLO NUBIAN BREED

2 DEVELOPMENT AND DISTRIBUTION OF THE ANGLO NUBIAN BREED

2.1 Description of the Anglo Nubian breed The Anglo Nubian goat is named after its origin from England and the long, droopy ears and convex nose associated with the name Nubian (Reinhardt and Hall, 1978). In the USA, the breed is usually spoken of as the Nubian. Herdbooks are maintained in USA, Britain, Canada and Australia (Mason, 2002). The Anglo Nubian is a tall goat with body weights of 50 to 70 kg in females and 60 to 80 kg in males (Peacock, 1996). The head is characterized by the high-arched Roman nose, ears are long and lopped; horns may be absent. The hair is short and silky. All colours occur: pure white to solid black, all shades of red and brown, multi-toned and spotted; white trim is common, particularly on the ears (Gall, 1996). Among dairy breeds developed in northern countries, Anglo Nubians are considered as a breed suitable for meat production because of its conformation and fertility, its adaptability to tropical conditions and non-seasonal breeding (Gall, 1996). Milk yield is less than in dairy breeds developed in Switzerland, but milk fat content is higher. Anglo Nubian goats registered by the American Dairy Goat Association (ADGA) (n=480) in 2003 averaged 826 kg milk with 4.9 % fat. A total of 1324 Nubian lactations recorded by the Dairy Herd Improvement Scheme (DHI) yielded 629 kg milk with 4.6% fat in 2002 (http://adga.org/). It was found that Anglo Nubian goats produce milk with high levels of total solids, protein and casein, as well as alphas1-casein. This has important implications on the value of Anglo Nubian milk for cheese production (Clark and Sherbon, 2000), especially under tropical conditions (Devendra, 1972). Additionally the value of Anglo Nubians as crossing partners for the development of breeds with favourable traits for cheese production have been proven (Dimassi et al., 2005). In Venezuela, Pariacote and Ruiz (2004) and Pariacote (1992) observed low reproductive performance with a tendency to decline with the size of the reproductive population in introduced breeds, among them the Anglo Nubian. As these breeds have been introduced for crossing with the Criollo population and have not been managed as purebreds the number of purebred males nowadays in most of the cases is less than 20, resulting in difficulties to preserve themselves as a population. Heinkel (1986) observed production potential of Criollo and crossbreds with Anglo Nubians on a farm in Venezuela; purebred female Anglo Nubians were not kept there because of low adaptability and consequently, low yields. Garcia et al. (1996) found abortion rates of 6.2%, 14.6%, and 14.9% in Criollo, halfbred Anglo Nubian and purebred Anglo Nubian, respectively. In a report from Colombia (Berrio et al., 1995), the milk yield in Anglo Nubians was less than in Criollos when both breeds received the same improved feeding and management.

2.2 Origin and breed development of the Anglo Nubian The Anglo Nubian breed was developed in Britain at the end of the 19th century (Peacock, 1996). Literature is not consistent as to which goat breeds contributed to the development of the Anglo Nubian. Porter (1996) states that the origins of the Anglo Nubian are diverse and were initially somewhat random. It originated from haphazard crosses between the native prick-eared Old English goats and a variety of lop-eared breeds from the eastern Mediterranean, north and east Africa and India, which had in common the carriage of their DEVELOPMENT AND DISTRIBUTION OF THE ANGLO NUBIAN BREED 369 ears and a roman facial profile to a lesser or greater degree. The imported animals were long- legged, hardy goats; well adapted to hot and dry climates. A little Swiss blood was included as well. According to Peacock (1996) only the two tropical breeds – Zaraibi from southern Egypt and Jamnapari from India – were crossed with the local British goat. Gall (1996), Mason (1981) and Mason (2002), name the lop-eared goats Zaraibi and Jamnapari, as well as local British and Swiss goats, but also another breed identified only by the name of its place of origin – Chitral – in the extreme north of Pakistan. The Jamnapari (also called Jamnapuri or Jumuna Pari) derived its name from the location of the breed beyond the river Jamna (Jamna Par) in Uttar Pradesh in the north of India (Gall, 1996). It is a dual-purpose milk and meat type much valued for its good milk yield from a large udder (Porter, 1996). It has been widely used for crossbreeding in India and elsewhere and has been used to upgrade goats in South-East Asia, particularly Indonesia, where it is known as the Ettawah breed (Gall, 1996; Peacock, 1996). The Jamnapari has contributed to form the Boer goat in South Africa and the Anglo Nubian. Nowadays, purebred Jamnapari are estimated at numbers less than 5,000 does (Gall, 1996). The name of the Zaraibi relates to confinement, indicating that this is the type of goat kept as a dairy animal under more intensive management in Egypt (Gall, 1996). It is now rare in Egypt raised mainly in the North Delta and around Cairo in small flocks as a household dairy animal. It is preserved on experimental farms of the Ministry of Agriculture. Mason (1981) states that there are no recent descriptions of the Zaraibi but Porter (1996) describes it as a breed used for milk production with a spherical udder traditionally protected by a leather bag. The Shami or Damascus breed of Syria is close to the Nubian type but its coat is long. It is a good milker (Mason, 1981). Gall (1996) states that the Damascus is one of the main dairy breeds of eastern Mediterranean countries and Iraq. It belongs to the Nubian group, together with the Zaraibi and Sudanese Nubian. Nubian goats and Indian dairy breeds (Jamnapari) may have a common ancestor in ancient Iran. It is the dairy goat of the urban areas in Iraq, Lebanon, Jordan and Syria. The Sudanese Nubian breed is a black goat of the Nubian type with grey ears, coarse and long hair. It is of medium size and has a typical Roman nose with its convex profile. Horns are nearly always present and it is known to be a good milker (Mason, 2001). Its chief breeding area is the south of the territory of ancient Nubia, but its appearance relates it to the Syrian Mountains. It forms the bulk of the goat population in Sudan. The Old English Goat (also called Common English Goat) was described by Pegler in the late nineteenth century, cited by Porter (1996) as a long-bodied, square-shaped goat with a neat, tapering head and prominent frontal bone. The ears were rather large, erect or horizontal and pointing forward. The coat was fairly close; there was often a fine, soft woolly undercoat. According to Gall (1996) and Mason (1981), the Old English goat has been completely absorbed in the crossbreeding with goats imported from Africa and Asia to form the Anglo Nubian. The Old English Goat was no longer recognized as a breed after 1939 (Reinhardt and Hall, 1978) and is now considered extinct (Mason, 2002). During the latter half of the nineteenth century, steamers continued the customary shipping practice of carrying on board goats acquired abroad for the homeward journey. On docking, these exotic goats from various sources were often bought by goat keepers and crossed with their own animals to increase size and milk yield. Many of these goats came from India, 370 DEVELOPMENT AND DISTRIBUTION OF THE ANGLO NUBIAN BREED especially from the Chitral region. “From the East” came Nubian, Egyptian, Assyrian and Syrian goats and the Zaraibi of Egypt. The term Nubian came to be applied to all these goats from the Near/Middle East as long as they had the typical features of height, arched profile and long lop ears. The term Anglo Nubian was given to the various crossbreds in the British Goats Society herdbook in 1893. For a while, the breeders began to lose interest in the “Nubian”; the type increasingly diluted its eastern look and was in need of fresh blood. That all changed from 1896 to 1904 when one Indian Jamnapari, one Zaraibi, one Chitral and one buck of unidentified breed were imported (Porter, 1996; Reinhardt and Hall, 1978). These four bucks were used in crossbreeding with the Old English Goat to produce the Anglo Nubian. A small percentage of Swiss blood was incorporated as well. The two Indo-Pakistan billy goats were particularly successful and, between them, these four fresh imports sired the first hundred or so Anglo Nubians. In 1910, the Anglo Nubian was recognized as a breed in Britain and registry began with 459 goats accepted as the nucleus of the Anglo Nubian section of the herdbook (Reinhardt and Hall, 1978). The early Anglo Nubian was expected to produce adequate milk with high fat content and to grow very large and fast; especially bucks were used for draft purposes. In 1911, some modifications were made in the original standards and over 60 more goats were allowed in the Anglo Nubian section. Breeders tried to import more animals in the following years, but due to health restrictions, no more imports to Britain were permitted (Reinhardt and Hall, 1978; Porter, 1996).

2.3 Distribution of the Anglo Nubian Main distribution flows are those starting from Britain to the USA and Canada and from there further on to Latin America (Figure 1). Where information was available, flows to Africa, the Middle East and Oceania are briefly outlined. Nubian type goats were imported into the USA as early as 1896, but most of them were either not disseminated or were not used on purebred animals so that they had no impact (Reinhardt and Hall 1978). In 1909 the importation of one buck and two does from Britain formed a nucleus from which Nubians in America descended. In total, about 30 Anglo Nubians were brought from Britain to the USA between 1909 and 1950 (Gall, 1996). Between 1909 and 1918, “Anglo” was dropped from the name, and 40 animals were registered as purebred Nubians. No new blood was ever incorporated in the Anglo Nubian of the USA (Reinhardt and Hall, 1978). In 1999 there were 1,224 registered Anglo Nubians (DHIA, 2000, cited by Gall, 2001). Anglo Nubians were imported from Britain to Canada in 1909, but no records were kept. In 1917, 5 more Anglo Nubians were imported, one of which later was sold to California. By 1921, a registered breeding program had been established in Canada. Offspring were imported into the USA; those purchases of Canadian imported and bred Anglo Nubians continued to have a great impact on Nubians in the USA until the late 1940’s (Reinhardt and Hall, 1978). Anglo Nubians were exported from Britain to the West Indies and thence to the USA after 1910 (Porter 1996). Anglo Nubians were exported from the USA to Puerto Rico and South America in the 1940’s (Reinhardt and Hall, 1978), where they have been used in goat development efforts. Crossbreeding programmes with Anglo Nubians from USA were established in Mexico, Venezuela and Brazil where importations of animals from other countries were virtually impossible because of animal health regulations (Gall, 2001). DEVELOPMENT AND DISTRIBUTION OF THE ANGLO NUBIAN BREED 371

Anglo Nubians were first imported to Argentina, early in the 20th century, and spread mostly in the northern half of the country. The diversity of the genetic pool, different environmental conditions in the different parts of the country, free mating and natural and artificial selection has led to the development of different local populations (Lanari et al., 2003). In the 1960’s Anglo Nubians from Canada and Britain were introduced to the province of Cordoba and Anglo Nubians from Brazil to the province of Santiago del Estero. A breeding nucleus was established in the central region of the country, in Villa de Maria del Rio Seco in the province of Cordoba, from where breeding males were distributed and intensively crossbred with the Criollo goat. Goat keepers were interested in crossbreeding as they considered that reproduction and weight gain were much higher than in the Criollo (Mueller, 1994). In Mexico, the central and northern states have Criollos which, according to the predominant types of external appearance, are predominantly Granadina and Nubian crosses (Montaldo et al., 1995). The widespread use for many years of Anglo Nubian, Alpine, Saanen and Toggenburg sires imported from the USA makes current Mexican Criollo goats a multi-breed population composed mainly of Nubian crosses in herds used for meat production, and crosses of Granadina with breeds of Alpine origin in milk producing herds (Montaldo and Meza, 2000). In Chile, Anglo Nubians were introduced from the USA since the 1970’s (Burrows, 1994). The Criollo goats of today seem to be in fact high grade Anglo Nubians according to the large size of the ears, and the lack of other long eared breeds in the country (Montaldo and Meza, 2000), a finding confirmed by Gallo and Wainnright (1995) who observed the form of the ears of goats in the south of Chile to be erect in only 4.5%, drooping of medium length in 67.5% and long lopped in 28.0%. In Cuba, importations of milk goats commenced in 1986, mainly of Anglo Nubian, Alpine, Saanen and Toggenburg. Cuban Criollos were crossed with these breeds mainly in an uncontrolled way. Today, there are few pure Criollos left, most of which are found in the East of the country. Criollos nowadays are mainly crossbreds of Anglo Nubian or Alpine descent (Ribas et al., 2000). Peru imported Anglo Nubians from the USA in 1961. Velez and Callacna (1984) used the descendents of these Anglo Nubians for a long-term study on performance of Criollo and crossbred goats. In Ecuador, Anglo Nubians were used in crossbreeding programmes (Gomez, 2003; Narvaez and Hernandez, 1995). A project located at the National University of Loja imported 2 bucks and 25 female Anglo Nubians from Peru in 1998. Bucks were rotated in flocks of varying degree of Criollo and Anglo Nubian crossbred blood; up to date the University of Loja keeps a flock of purebred Anglo Nubian (Mendoza, 2004). In Brazil, the first importation of Anglo Nubians from England took place in 1929 by private entrepreneurs. Other importations in 1932 from the USA and England introduced Anglo Nubians to Bahia from where they spread to the Sertao region. From 1938 onwards, governmental agencies became active in the introduction of Anglo Nubians. In those years, Brazilians experienced a period called "the empire of the ears" referring to zebuine bovines whose large, pendulous ears were seen as a sign of success; that's why goat keepers got enthusiastic about the arrival of a long-eared goat breed. In the last years, a lot of importations of Anglo Nubians occurred, mainly from the USA. The breed is widely used in the Northeast of Brazil in crossbreeding for meat and milk. It is estimated that from a total of 372 DEVELOPMENT AND DISTRIBUTION OF THE ANGLO NUBIAN BREED

10 millions of goats, about 7 millions have some influence of Anglo Nubians (personal communication M. N. Ribeiro, 2005). In Oman, Anglo Nubians were imported from Britain in 1988 to improve indigenous goat performance; the project is still under evaluation but there are signs of inbreeding and lack of interest among farmers (Zaibet et al., 2004). In Thailand, Anglo Nubians were used in a crossbreeding programme (Sripongpun, 1991). Original imports to Kenya were , Saanen, Anglo Nubian and Toggenburg; mainly Saanen under a Dutch aid project in the 1960-1970’s. Respiratory diseases prevailing in improved breeds have been a major reason for restriction of their impact and distribution. In the Ethiopian highlands, a dairy goat development programme implemented crossbreeding and improved goat management between 1989 and 1997. Anglo Nubian goats were introduced from Britain and crossed with the local Somali goat (Peacock, 1996), A year after the programme had finished, Ayalew (2000) studied the impacts of the project and found that the attributes of crossbreds were not maintained because the pool was too small to maintain 50% exotic blood level in the crossbreds, which ranged from 6.25 to 75%, with the 50% crosses representing less than a quarter of the crossbred population. Shortages of crossbred breeding males also led to gradual backcrossing of the does, resulting in an increasing mosaic mix of crossbreds. In Egypt, Anglo Nubians are used on intensively managed farms to produce French type cheese; some of these farms face serious financial troubles and technical problems like adaptation of imported animals and securing replacements (Galal, 1995). In New Zealand the Anglo Nubian together with British Saanen and British Toggenburg were used to crossbreed with selected goats of the feral goat population of that country. Crossbreeding and interbreeding at the F2 and F3 generations resulted in a meat type synthetic breed named Kiko (Mowlem, 1992). In Ethiopia, Kenya (Peacock, 1996; Rewe et al., 2002), Ecuador (Gomez, 2003; Narvaez and Hernandez, 1995) and Thailand (Sripongpun, 1991), Anglo Nubians were reported to be used in crossbreeding programmes, but once the official crossbreeding projects expire, it is often not known how many purebreds or crossbreds remain. In most cases the optimum level of exotic blood required for efficient production and adaptation was not established (e.g. Kenya) and the attributes of crossbreeding were not maintained because the pool of breeding males was too small (e.g. Ethiopia, Oman). In other cases, crossbreeding resulted in higher production and better reproductive performance (e.g. Argentina). In some countries of Latin America, crosses with Anglo Nubians (and other specialized breeds) tend to cause the decline or even disappearance of the original Criollo goat. DEVELOPMENT AND DISTRIBUTION OF THE ANGLO NUBIAN BREED 373

Figure 1: Gene flow of the Anglo Nubian goat

Canada Great Britain High volume

USA Europe Low volume

Mexico

Belice North and Central Israel America Middle East Oman Asia

British Caribbean India Banglade Venezuela

Guayana Malasia Ecuador

Perú Brasil Australia Oceania Bolivia Chile Ethiopia New Zealand Caribbean and Argentina South America Kenya

Africa

Less frequently Anglo Nubians in tropical countries are used as purebreds. Today there are purebred Anglo Nubian populations in Mexico, Brazil, Peru, Honduras, Trinidad, Egypt, Oman, India, Bangladesh, The Philippines, Mauritius, Malaysia, Israel (Gall, 1996), and Colombia (SENA, 1991). Purebred Anglo Nubians in Colombia yielded less than Criollos when both breeds received the same improved feeding and management. But in Israel, Anglo Nubian goats descendants from an import of bucks and frozen semen from the USA in the 1980’s adapted well to the Mediterranean environment (Landau et al., 1995). In the British Caribbean, purebred Anglo Nubians are kept on private as well as governmental farms and are used in crossbreeding; they have been more successful than the British Alpine, Toggenburg and Saanen breeds. Anglo Nubians have also been introduced into Belize, Grenada, St. Lucia, St. Vincent, Antigua, Trinidad and Tobago and Guyana (Devendra and Chenost, 1973).

2.4 Conclusions i The Anglo Nubian breed was developed in Great Britain by using local and imported goats. Of the breeds used, the British base is now extinct (Old English Goat). Information is scarce on numbers of the tropical and subtropical foundation stock (Zaraibi), the number is alarmingly low in the Jamnapari as it is being used extensively in crossbreeding programmes both in its country of origin (India) and in other developing countries. ii The Anglo Nubian breed spread to all continents. Apart from being kept as purebreds, it is more often used in crossbreeding programmes in different regions of the world. 374 DEVELOPMENT AND DISTRIBUTION OF THE ANGLO NUBIAN BREED iii The value of this genetic resource has been recognized a long time ago, but there seem to be no efforts to counteract the danger of loosing it by excessive use in crossbreeding. THE KEEPING OF ANGLO NUBIAN GOATS IN BOLIVIA: HISTORY, 375

3 THE KEEPING OF ANGLO NUBIAN GOATS IN BOLIVIA: HISTORY, DEVELOPMENT AND IMPACTS

3.1 Conditions of goat keeping in Bolivia Bolivia is a country with very diverse climate and geography, ranging from tropical rain forests to the high peaks of the Andes. Goats are mainly kept in the inter-Andean valleys at altitudes ranging from 1,000 to 3,000 m and in the Chaco region at altitudes ranging from 150 to 1,500 m. In these zones, annual rainfall is on average 400 to 500 mm, distributed mainly during the months of December to March. According to estimates, there are about 1.5 million goats in the country (CID, 1996; FAO, 2003), 92% of which are kept in the departments of Potosi, Chuquisaca, Tarija and Cochabamba. The vast majority of goats are of the Criollo breed, known for its hardiness and adaptation to seasonal food and water shortage. Its products in order of importance to the smallholder farmers are: manure, milk, meat, and hides. Other uses of goats are important too, as, for example, a form of insurance, banking reserve and means of paying social obligations. Most of the goats are kept in mixed flocks with sheep. These flocks are indivisible components of the local smallholder production systems (Iñiguez, 2004). The majority of goat holders are subsistence oriented and less than half of the milk and meat produced enters the market (Ayala, 2002; Sanabria et al., 1992; IICA, 1989). Especially in regions where the keeping of cows is limited, goats are the main producers of milk for the smallholder family.

3.2 Introduction of Anglo Nubian goats to Bolivia There are no official records on the introduction of goats from outside Bolivia. The following information is a compilation of project reports, personal, telephone or e-mail interviews with development workers of governmental and non-governmental organizations, goat keepers and other experts. The first importations of Anglo Nubian goats to Bolivia took place in the late 1960’s (Chumacero, 1993, cited by Iñiguez, 2004). Since then, there have also been exchanges between different parts of the country. The introductions took place in the regions with a tradition in goat keeping, namely the departments of Cochabamba, Tarija, Chuquisaca, Potosi, and some parts of Santa Cruz. Table 1 summarizes the data available on the introduction of Anglo Nubians to Bolivia. In 1985, 24 Anglo Nubians (19 females and 5 males) were imported from Argentina to the department of Cochabamba. These animals were kept in the instalations of a newly founded Goat Project at the Faculty of Agronomy, University Mayor San Simon in Cochabamba. The flock at the faculty served as a multiplier in order to introduce young male goats to Criollo flocks in the province of Esteban Arze, department of Cochabamba. The field work was conducted by the Non-Governmental Organization (NGO) Sociedad Nueva (SONU) in cooperation with the Faculty of Agronomy (Comité de Coordinación Interinstitucional, 1987). This work was concluded in 1991. Since 1994, the goat project at the faculty cooperates with German counterparts, and deep frozen semen of Anglo Nubian bucks was imported from Germany in the late 1990s. 376 THE KEEPING OF ANGLO NUBIAN GOATS IN BOLIVIA: HISTORY,

In the interandean valley of Tarija, the Association of Goat Breeders (ACCT, Asociación de Criadores de Cabras Tarija) imported 38 Anglo Nubians (3 males and 35 females) from Santiago del Estero, Argentina in December 1996. These animals were kept in the central valley of Tarija and used on the one hand for purebreeding to maintain a small flock of purebreds, and on the other hand for crossbreeding with Criollo goats owned by the members of ACCT. This project was financed by the European Union and the World Food Programme and carried out by PRODIZAVAT (Programa de Desarrollo Integral de la Zona Andina y Valles Altos de Tarija, Integrated Development Programme for the Andean Zone and High Valleys of Tarija) (personal communication M. Bass Werner, 2004). Again in Tarija, a private farm not far from the city is managed intensively with purebred Anglo Nubians. The original stock was imported from Catamarca, Argentina. During the last years, the owner imported deep frozen semen from France in order to inseminate his goats (personal communication M. Bass Werner, 2004). In the Chaco region (which covers part of the Tarija, Chuquisaca and Santa Cruz departments), Anglo Nubians were introduced in the 1970s by individual goat owners, some of whom lived in the cities but owned large flocks of goats in the country side; others were goat keepers making a living of their goats and other livestock (personal communication J.L. Vaca, 2004). In the tropical lowland of Santa Cruz, the NGO Heifer International imported about 30 Anglo Nubians from the USA in 1984, but all of them died while in quarantine. Another importation of 20 females and 10 males took place in 1986. These animals were taken to the village of San Julian and donated to smallholder farmers (personal communication, R. Hinojosa, 2005). A private farm in Santa Cruz imported the first Anglo Nubian bucks in 1975 from Paraguay. Later on they received bucks from the Heifer project. In the late 1990s, they bought male and female Anglo Nubians from Caritas, introduced via the Netherlands. In the last years up to 2004, bucks were bought from private persons who imported Anglo Nubians from Brazil (personal communication, C. Foyanini, 2005). The University of Santa Cruz in 1986 used 100 doses of deep frozen semen of Anglo Nubian and French Alpine breeds for crossbreeding with local Criollo goats in order to improve milk production (Videz and Cardona, 1992); the report does not specify the number of doses according to breed, nor are there other reports of the results of this project. There is a demand for purebred Anglo Nubians in Bolivia, which is not always covered by the number of animals offered for sale. In the south of the country, people resort to buying goats in Argentina. Apart from that, potential buyers and sellers sometimes meet with difficulties to get together simply because they don’t know where to get relevant information.

3.3 Spread of Anglo Nubians in Bolivia After the introduction of the Anglo Nubian to Bolivia it spread within the country through the promotion by various organisations. Table 1 summarizes the data available on the spread within the country. THE KEEPING OF ANGLO NUBIAN GOATS IN BOLIVIA: HISTORY, 377

Table 1: Introduction of Anglo Nubian goats to Bolivia and spread within the country

Place of introduction Year Number of Place of origin males females Cochabamba, Faculty of 1985 5 19 Argentina Agronomy 1990’s semen Germany Cochabamba and 1989-92 14 Cochabamba, Province Santa Cruz Campero 2000 67 Santa Cruz Cochabamba, Province 1990’s 17 Santa Cruz Mizque 1996 3 18 Santa Cruz Tarija 1996 3 35 Argentina Chaco 1970’s - - Santa Cruz 1975 - Paraguay 1984 30 USA 1986 10 20 USA 1986 semen - up to2004 - Brazil 1994 3 9 Santa Cruz Potosi 1994 2 Cochabamba 2000 10 60 Santa Cruz Chuquisaca 2000 11 Tarija -: data not available In the 1990’s the Governmental Agronomic Research Institute IBTA (Instituto Boliviano de Tecnologia Agropecuaria) bought 17 Anglo Nubian goats in Santa Cruz to be kept at the premises of IBTA in the province of Mizque, Cochabamba and in four individual intensively managed flocks, also in the valley of Mizque (Campero, 1996; personal communication R. Ergueta, 2004). From October 1989 to May 1992, the PDAR (Programa de Desarrollo Alternativo Regional, regional programme for alternative development) worked in the valleys of Mizque and Campero. The programme introduced 14 Anglo Nubian bucks to the region, which were bought from the Faculty of Agronomy in Cochabamba and from individual owners in Yapacani, Santa Cruz (Caballero, 1994). In 1996, 3 bucks and 18 females were bought in Santa Cruz and introduced to Mizque, where they were kept at the installations of the NGO CEDEAGRO in Bolivia (Centro de Desarrollo Agropecuario, Centre for Agronomical Development). Occasionally, a male goat was lend out to goat keepers (Stemmer, 1996). In November of 2000, the World Food Programme (WFP) introduced 67 Anglo Nubians bought mainly in Santa Cruz to the province of Campero, Cochabamba. All the animals were given directly to the members of an association of goat breeders and some to people who had not owned goats before (Stemmer, 2003). 378 THE KEEPING OF ANGLO NUBIAN GOATS IN BOLIVIA: HISTORY,

In Potosi, two Anglo Nubian bucks and nine females were introduced from Santa Cruz in 1994 and kept at the IBTA research station in Chinoli (personal communication E. Chumacero, 2004). In 1994 a further two bucks were purchased from the flock at the Faculty of Agronomy in Cochabamba (Stemmer, 1994). In Chuquisaca, the World Food Programme financed a goat project, which in 2000 introduced 70 Anglo Nubians from Santa Cruz (10 males and 60 females) and eleven from Tarija (all of them bucks). These animals were given directly to the members of an association of goat breeders in the village of Potreros (PMA-DRIPAD, 2003).

3.4 Development of Anglo Nubian populations in Bolivia The following paragraphs show the development of the Anglo Nubian population in Bolivia by region. Table 2 summarises the findings.

3.4.1 Cochabamba The goat project at the Faculty of Agronomy in Cochabamba is currently maintaining a goat flock of some 25 female Anglo Nubians and some 15 crossbreds with Criollos (F1 and backcrosses to Criollo and Anglo Nubian) and a few pure Criollos. The project sells young pure and crossbred bucks to development agencies and private goat keepers. All the goats are kept in one flock and managed the same way, foraging natural pasture from December to June and receiving hay and maize silage from July to November. Performance of the different genotypes was evaluated in various investigations, some of which are presented here as there is no information from other parts of Bolivia about the performance of pure Anglo Nubians, Criollos and their crossbreds kept under the same conditions. In the flock of the faculty, pure Anglo Nubians showed higher live weights than F1- crossbreds and Criollos from birth up to 4 months of age, while there was no difference between the latter two. Mortality from birth to one year of age was much higher in Anglo Nubians (15.0%) than in F1 (7.0%) and Criollos (3.4%) (Condori, 2000). Milk production was significantly higher in Anglo Nubians than in Criollos; this superiority was confirmed when performance was compared on a metabolic weight basis (Pari, 1998). Up to now, the project has shown sustainability, managing to auto-finance its nucleus flock during the last 5 years. In the province of Esteban Arze, the introduced pure male Anglo Nubians were kept in a centre built in the village of Julo Chico and mated to Criollo females brought from the surrounding countryside to the centre. There was some extension service and aid for improving night enclosures. The project worked for 3 years (from 1987 to 1990) and after that, the animals and installations of the centre were transferred to the community. Nowadays, there is just one family who keeps crossbreds in semi-confinement and a few others who maintain some crossbreds in their flocks. However, the majority sold or slaughtered the purebreds and crossbreds (personal communication F. Cautin, 2004). The reasons for the practical disappearance of the Anglo Nubians can be found in the poor access to a market for goat cheese and, probably, in extension work that was not focused on the real constraints of goat husbandry; this conclusion is drawn from the analysis of a booklet published by SONU and aimed at the goat keepers (Mosua, 1989), which does not address solutions to the problems of goat keeping in an understandable way and got some outright THE KEEPING OF ANGLO NUBIAN GOATS IN BOLIVIA: HISTORY, 379 mistakes in it. In contrast it is worthy to mention the work of WFP in Tarija (see below), where a useful and easy to understand booklet was published (Pinaya, 2002). In the province of Mizque, the Anglo Nubians kept at the research station of IBTA were managed intensively and used in feeding experiments. At the closure of the station in 1998, these animals were transferred to the NGO CEDEAGRO. This NGO already had a flock of goats; of the 21 animals originally bought in Santa Cruz, approximately 60% died in the months following the transfer from the tropical lowlands to the inter-Andean valley of Mizque at 2,000 m altitude. The kids born from the remaining females, though, adapted well. CEDEAGRO sold Anglo Nubians and crossbreds to interested goat keepers and offered extension services (Eisele, 2001; personal communication, R. Ergueta, 2004). In 2001, these activities came to an end when CEDEAGRO closed down its agronomic projects for lack of outside funding. Five purebred Anglo Nubians (1 buck and 4 females) were bought by the Agricultural Faculty of Cochabamba and incorporated in its nucleus flock (Stemmer, 2001). Nowadays, there are flocks, which maintain crossbred goats, mainly near the villages or close to the interdepartmental road. At these locations, goat keepers sell soft cheese produced from goat’s milk. In more remote places, the amount of cheese produced is less and more than half of it is consumed by the family itself. In the province of Campero, 67 Anglo Nubians were introduced from Santa Cruz to the community of Novillero in November 2000. They had mortalities of 54% in the four months following the transfer; only one buck and 30 females survived. This can be attributed to the lack of extension, lack of sanitary care, and lack of experience in the case of the people who received Anglo Nubians and had not owned goats before (Stemmer, 2003). Up to December 2002, only fifteen purebred kids were born in Novillero; most of the purebred females crossed with Criollo bucks (personal communication O. Hinojosa, 2003). The WFP has ceased to assist goat keepers in Novillero at the end of the year 2002. Nowadays, purebred and crossbred Anglo Nubian are still kept by some goat keepers. The Anglo Nubian bucks introduced to the province of Campero by PDAR had 50% mortality. Primiparous Criollo goats suffered an increase in distocic parturitions due to bigger size of crossbred kids. Growth of these kids was impaired as the Criollo dams did not provide sufficient quantity of milk (Caballero, 1994). The report does not mention if there was an extension service working with the goat holders.

3.4.2 Chaco In the part of the Chaco region that belongs to the department of Santa Cruz, the population of Anglo Nubians did not thrive well. As the flocks there are managed extensively, let out to graze on shrubs and thorn bush, the main problem were the large udders of the females, which frequently got entangled and hurt by the thorns resulting in high levels of mastitis. Another problem was the supply of breeding animals, which could not be sustained. Nowadays, there are no purebreds in the Chaco and very few goats remain with some characteristics of the Anglo Nubian like the pendulous ears (personal communication J.L. Vaca, 2004). A study carried out in the province of Gran Chaco, department of Tarija, including 210 flocks found that 86% of goats are Criollo and 14% are crossbreds with some characteristics 380 THE KEEPING OF ANGLO NUBIAN GOATS IN BOLIVIA: HISTORY, of Anglo Nubian or Toggenburg; no specification as to the proportion of the two introduced breeds are given (CODETAR, 1992).

3.4.3 Tarija In Tarija, the members of ACCT produce goat cheese and manage their goats on natural pasture with very little additional production of forage. With aid of the WFP, the management of the goats was improved by better housing (construction of roofs in the night enclosures), veterinary service, pasture enclosures and rotations (Stemmer, 2003). The introduction of Anglo Nubians was done step by step by crossbreeding with female Criollos. The F1 and ¾ backcrosses to Anglo Nubian have satisfactory levels of milk production and adaptation, and ACCT has decided to keep crossbreds and pure Criollos in order not to lose the high adaptation of the latter. Another reason for keeping Criollos is the considerable rise in milk and meat production resulting from an improvement of feeding and management such as given to the crossbreds (personal communication M. Bass Werner, 2004). The private farm in the central valley of Tarija produces four different French types of goat cheeses, which are sold in different cities of Bolivia. The goats are kept in complete confinement and are stall fed. The size of the flock is small and has been stable for some years (personal communication M. Bass Werner, 2004).

3.4.4 Chuquisaca The Anglo Nubians introduced by WFP in Chuquisaca in 2000 suffered very high mortality (53%) and abortion rates (70%). Of the 81 purebreds introduced, 38 were still alive in 2003. The reasons for this sharp decline were identified as very limited extension service, lack of veterinary care, deficient design of night enclosures, which were closed buildings with small doors prohibiting sufficient air circulation and restricting animals’ movements so that there was more fighting among animals and consequently, a rise in abortion rate. Feeding was deficient as no additional fodder was produced and the region suffers from water shortage in the dry season. At the same time, cheese production of goat milk was only sporadic, so that there was no direct financial incentive connected with the keeping of Anglo Nubians (Stemmer, 2003).

3.4.5 Potosi The Chinoli research station lent Anglo Nubian bucks to smallholder goat keepers, which preferred this breed over the Saanen for its coloured hair coat and long ears. At the station, crossbreds of Anglo Nubian and Criollo reached higher body weights than crossbreds of Saanen and Criollo. According to the technician responsible for the goat programme, E. Chumacero (personal communication, 2004), the Anglo Nubian purebreds did not adapt well to the altitude of Chinoli (3,450 m altitude). At the closure of the research station in 1996, all the goats were sold and there are no reports on their fate. According to Yagil (1991), the introduction of Anglo Nubian bucks was not successful. High mortality of purebreds and low productivity of crossbreds were due to lack of extension service so that goat farmers were left on their own and could not even envisage short-term advantages and so left the breeding regime. Adaptive problems of the imported breed were not taken into account, so lack of improvement in management and no sanitary attention depreciated genetic productivity capabilities. THE KEEPING OF ANGLO NUBIAN GOATS IN BOLIVIA: HISTORY, 381

3.4.6 Santa Cruz The introduction of Anglo Nubians to San Julian in Santa Cruz was first met with enthusiasm as the animals adapted well and showed high prolificacy. Heifer International set up an extension service, first on animal health, management and milk production aspects, which later on was reduced to vaccination and parasite control programmes. The animals were found to be very susceptible to internal parasites. Passing the years, the interest in keeping purebred Anglo Nubians faded and crossbreeding with Criollo goats occurred more often, being the main reason the lack of interest in producing milk, according to R. Hinojosa (personal comunication, 2005). As the Anglo Nubians were not milked there was no benefit of keeping milk goats. Heifer International did not offer help in the marketing of milk or milk products, but there were other institutions in the region, which did, like the Lutheran Church who constructed a milk-processing unit. This plant, however, never got to function (personal comunication R. Hinojosa, 2005). Purebred Anglo Nubians were sold to private breeders in Santa Cruz, to the NGO Sociedad Nueva of Cochabamba and the IBTA research station Chinoli in Potosi (personal comunication R. Hinojosa, 2005). The Anglo Nubians introduced to a private farm in Santa Cruz adapted very well and had low mortality; the only problem were the higher nutritional needs compared to Criollo goats. Nowadays the family manages two flocks of purebred Anglo Nubian goats, one of 500 animals in Pailon at 50 km from the city of Santa Cruz and another one of 120 animals in San Javier. The latter location poses more problems to goat keeping due to higher humidity and parasite burden. The goats are managed semi-intensively with supplementary feeding during the dry season and parasite control. Milk and cheese are produced; the marketing of milk poses problems due to lack of time to dedicate to the sale. Breeding animals are sold mainly to NGOs and to private persons. Table 2 summarizes the development of Anglo Nubian populations in Bolivia by place of introduction. 382 THE KEEPING OF ANGLO NUBIAN GOATS IN BOLIVIA: HISTORY,

Table 2: Development of Anglo Nubian populations in Bolivia

Place Number Mortality Actual number of System of Market Extension of intro- of intro- purebreds crossbreds management access service duced duced goats goats (%) (year) Cochabamba Faculty 24 - 25 15 semi- good offers Cochabamba (1985) intensive extension E. Arze - - 0 few extensive poor poor (1987- 1990) Mizque 17 - 0 * - intensive good good (IBTA) (1990’s) Mizque 21 app. 60 - - semi- good good (1996) extensive Campero 14 50 - - extensive difficult - (1989- 1992) Campero 67 54 few few extensive difficult sporadic (2000) Chaco Unspecified - - 0 very few extensive poor none (1970’s) Tarija ACCT 38 low app. 60 app. 400 semi- good good (1996) intensive private farm few - stable 0 intensive good - (1990’s) Chuquisaca Potreros 81 53 38 few extensive poor poor (2000) Potosi IBTA Chinoli 13 - 0 * - intensive difficult - (1994) Santa Cruz In quarantine 30 100 0 0 - - - (1984) San Julian 30 3 few - Semi- poor good in the (1986) extensive begin- ning, later reduced Pailon and - low 620 0 Semi- good - San Javier (1975 - intensive 2004) -: data not available; *: all sold at closure of station; ACCT: Asociación de Criadores de Cabras, Tarija; IBTA: Instituto Boliviano de Tecnologia Agropecuaria THE KEEPING OF ANGLO NUBIAN GOATS IN BOLIVIA: HISTORY, 383

3.4.7 Comparison with other exotic goat breeds Compared to other introduced goat breeds, the Anglo Nubian was the most persistent. Importations of Saanen, Alpine and Toggenburg to Bolivia took place in the 1980’s and 1990’s, but very few crossbred animals of these breeds can be found nowadays and, to the authors’ knowledge, no purebred animals. The Saanen was very susceptible to high radiation (personal communications, R. Ergueta, 2004 and E. Chumacero, 2004), while specific reasons for the disappearance of the Alpine and Toggenburg are not known to the authors. In Argentina, a similar situation occurred regarding the importations of Saanen, Toggenburg and Alpine goats, which were introduced without success. Crossbreds with Saanen and Criollo were able to maintain the adaptability of the Criollo but there was little rise in milk production and no benefit in meat production. Crossbreeding with Toggenburg resulted in a loss of adaptability without improving milk yield (Mueller, 1994); no specific reasons for the failure of crossbreeding with Alpine are given. In India and Venezuela, the Anglo Nubian was inferior to the French Alpine in milk yield. However, in these tropical conditions, the Anglo Nubian shows superiority over other exotics for reproductive traits. The local breeds are generally superior to all exotics for reproduction and viability. Circumstantial evidence from the field in Malaysia, Fiji, Mauritius and the West Indies suggests that breeds like the British Alpine, Toggenburg and Saanen are prone to reproductive failure and high mortality compared with the Anglo Nubian or Jamnapari (Quartermain, 1983).

3.5 Impacts of the keeping of Anglo Nubians

3.5.1 Impact on goat holders A very important advantage for the smallholders of keeping goats is the fact that these animals need very little monetary input while procuring a monetary income at a substantial profit margin (Ayala et al., 2004). The same situation was described by Primov (1994) for smallholders in Brazil where the author found that the overriding reason for the popularity of goats as the most profitable production line is the simplicity and economy, with which they can be produced. Fewer inputs mean that the prices received for goats contain higher profit margins. A constriction in the profit received from the sales of goats as a consequence of increased capital inputs will rapidly result in a deterioration of the relative advantages of goats to the small producer. This situation is likely to occur with the introduction of purebreds; however, the keeping of more adaptable crossbred goats seems to offer advantages to those smallholders who are able to market part of their products. In the case of the very few intensively managed goat enterprises in Bolivia, the keeping of pure Anglo Nubians is an advantage in the production of milk and milk products; on one hand, milk and fat quantity produced are high and on the other hand, the Anglo Nubian breed adapts well to the climate of the inter-Andean valleys if and when they are given the necessary care. It can be concluded that the introduction of purebred Anglo Nubians brings no benefit to small holders as the animals are likely to suffer high mortality and abortion rates. In the case that the animals were given to goat keepers as development aid, at least there was no direct monetary loss for the people involved. In San Julian, Santa Cruz, where purebreds adapted 384 THE KEEPING OF ANGLO NUBIAN GOATS IN BOLIVIA: HISTORY, better, goat keepers had some additional income from the sale of offspring (personal comunication R. Hinojosa, 2005). The introduction of Anglo Nubian crossbreds can result in a better position of the goat keeper. In a study carried out in Mizque, department of Cochabamba, daily milk yield of 62 Anglo Nubian crossbreds and 102 Criollo goats during the rainy season was 552 and 424 g, respectively; this difference was significant. During the dry season, daily milk yield of 14 purebred Anglo Nubians, 20 Anglo Nubian crossbreds and 85 Criollo goats was 692, 430 and 234 g, respectively (because of low numbers, no analysis of variance was carried out) (Altug, 2002). The higher milk yield of crossbreds can thus be expected in the rainy as well as in the dry season, a fact quite important for the successful raising of kids and for people who often don’t consume milk from other animals than goats. The same study (Altug, 2002) found no significant difference in litter size between 41 Anglo Nubian crossbreds and 67 Criollo goats; if this could be confirmed with a higher number of observations, there would be no loss to the goat holder in terms of kids born by preferring Anglo Nubian crossbreds over Criollo goats. In another study conducted in Mizque on 10 smallholder farms, five with flocks of Anglo Nubian crossbreds, and five with Criollo goats (Eisele, 2001), the following benefits and disadvantages of keeping Anglo Nubian crossbreds could be identified: In the flocks of crossbred goats, milk yield was higher than in Criollo flocks; in the latter case, goat holders frequently sold kids in order to obtain more milk for home consumption or for sale. The holders of crossbred goats did not need to resort to the sales of kids. Body weights of 19 crossbred and 64 Criollo goats, all adults, did not differ greatly. If this finding could be proved in a larger number of animals, the conclusion could be drawn that there is a positive impact of Anglo Nubian crossbreds as they produce more milk at the same body weight than the Criollos and, presumably, with similar feed intake. However, crossbreds may grow to higher body weights than Criollos with better feed supply. A possible negative impact is the greater need of the crossbreds for improved management, which is basically improved parasite control, i.e. a need for higher inputs. Eisele (2001) found the same level of ectoparasite infestations in crossbreds and Criollos although the former were subjected to better control measures.

3.5.2 Impact on the Criollo goat population Apparently, no marginalization or substitution of Criollo goat populations has taken place in Bolivia; but it must be stated that there is no way to quantify this impression of the authors, as to begin with, there is no reliable census of goat numbers in the country and the published figure of 1.4 million goats could be much higher (CID, 1996). Another limitation is the lack of distinction of breeds in the census. Taking into account the number of imported goats and their high mortalities in many cases, there is reason to believe that there are not more than a few hundred purebred and crossbred Anglo Nubians in Bolivia. A negative impact on the Criollo goat population, though, can be seen in the efforts and financing that was spent in introducing Anglo Nubians to climatic zones and management systems not suitable to the breed. Had the same effort been spent in characterizing the production environment and the genetic resource of the Criollo goat or in their improved breeding and management, our knowledge in these fields today would be far greater. THE KEEPING OF ANGLO NUBIAN GOATS IN BOLIVIA: HISTORY, 385

3.5.3 Impact on the environment In Tarija, there was a positive impact on the environment caused by the introduction of Anglo Nubians as there was a simultaneous improvement of the pastures caused by rotating flocks and the building of enclosures; plant biomass production in the enclosures is ten times higher than it used to be in the same area before it was fenced off (Stemmer, 2003). Contrary to this, the traditional management of communal grazing lands without any fencing practised by practically all Criollo goat keepers is detrimental to the environment, causing high grazing pressure, plant degradation, loss of species diversity, and, in some cases, soil erosion (Iñiguez, 1989; Sanabria et al., 1992; Caballero, 1994; Campero, 1996). 386 CONCLUSIONS

3.6 Conclusions Anglo Nubians were introduced to different production systems and climatic zones; successes or failures of these introductions can be summarized as follows: i Anglo Nubian populations increased in the cases of intensive or semi-intensive management including health care, improved housing and fodder production. These conditions are found in institutional flocks or those managed by private entrepreneurs; examples are the flock of the Faculty in Cochabamba, the ACCT and a private flock, both in Tarija. In these intensively managed goat farms, purebred Anglo Nubians have positive impacts because of their good adaptation to the climate of the inter-Andean valleys and their high yields. ii In semi-extensive management, this breed is able to survive but its production potential is not fully made use of. There are only few examples (flocks in Mizque) of persistent Anglo Nubian populations in these conditions, and all of them include some access to the market so that goat owners can sell cheeses at least during part of the year. Another prerequisite is the extension service needed to assist goat owners with the introduced breed so that they can respond to the higher needs of Anglo Nubians compared to the Criollo goats in terms of better housing, parasite control and improved feeding. iii In extensive management, Anglo Nubians did not last for long; cases include those in the province of Esteban Arze and Campero in Cochabamba, the Chaco, as well as flocks in Chuquisaca and Potosi. iv The introduction of purebred Anglo Nubians has brought no benefit to smallholders in Bolivia as the animals are likely to suffer high mortality and abortion rates. In the case that the animals were given to goat keepers as development aid, no direct monetary loss for the farmers was recorded. v In the case of introduction of purebreds, the impact of Anglo Nubians on smallholders is negative because the animals suffer high mortality and abortion rates. However, the introduction of crossbreds can result in benefits to those smallholders who are able to make use of the higher milk production of crossbreds by marketing part of the cheeses produced. A possible negative impact is the greater need of the crossbreds for improved management, which translates into the necessity of higher monetary input. vi The time needed for adaptation to a new environment can be very long when adult animals are transferred from one climatic zone to another, in some cases only the offspring of the introduced animals adapted well, as observed in Mizque. Purebreds cannot be transferred to the conditions of smallholdings without mortalities surpassing 50% and high abortion rates. When introduction is in the form of a slow process using young pure or halfbred bucks for breeding, results are much more favourable, like in Tarija. vii The impact of the introduction of Anglo Nubians on the Criollo goat population has been very limited. Today, Criollo goats far outnumber Anglo Nubians and crossbreds. A negative impact on the Criollo goat population, though, can be seen in the efforts and financing that was spent in introducing Anglo Nubians to climatic zones and management systems not suitable to the breed. Had the same effort been spent in characterizing the production environment and the genetic resource of the Criollo goat or in their improved breeding and management, our knowledge in these fields today would be far greater. CONCLUSIONS 387 viii The impact of the introduction of Anglo Nubians on the environment was positive in those cases where there was a simultaneous improvement of the pastures caused by rotating flocks and the building of enclosures. Inputs in infrastructure and management were responded to by high performance of the improved breed, making sense of restricting flock sizes and movements of the animals. ix The gene flow between developed countries was realized by private persons and on a commercial basis, whereas the transfers from developed to developing countries often involved development agencies. In the case of Bolivia, the gene flow was planned mainly by agencies, and only in few instances by goat keepers. There were also gene flows between countries of South America. x In Bolivia, the introduction of Anglo Nubians was met with failure when the planning was done merely by development agencies without considering or even knowing the production systems involved. On the other hand, careful planning and involvement of the goat keepers themselves led to success. xi The financing of the introduction of Anglo Nubians was mostly by development agencies, governmental or non-governmental. Only in few cases did private financing occur. It is concluded that in most cases, smallholder goat keepers did not suffer financial losses due to the introduction of Anglo Nubians, however public incentives frequently did not pay back. xii A negative impact of the transfer of Anglo Nubians lies in the fact that much effort was spent in introducing and adapting a foreign breed to the conditions of Bolivia, while the local genetic resource, the Criollo goat, has been neglected both by researchers and development agencies. The Criollo has, however, a high variability and production potential that can be utilised if these animals are included in research on management and breeding improvement (Altug et al., 2000). 388 REFERENCES

4 REFERENCES Altug, T., Valle Zárate, A., Stemmer, A., Ergueta, R. 2000. Phenotypic and production characteristics of Criollo goats in the interandean valleys of Bolivia. 7th International Conference on Goats, 15-21 May 2000, France. Ayala, V.J.L. 2002: Sistemas de produccion caprina en Mizque y Carrasco, departamento de Cochabamba. Graduate Thesis, Faculty of Agronomy, Cochabamba, Bolivia Ayala, V.J.L.; A. Stemmer; R. Garcia; R.Ergueta; A. Valle Zarate; J. San Roman 2004: Sistemas de poduccion caprina en Mizque y Carrasco, departamento de Cochabamba. Memorias de la XV Reunion de ABOPA, 24-26.11.2004, Oruro, Bolivia Ayalew, K.W. 2000: Do smallholder farmers benefit more from crossbred (Somali x Anglo- Nubian) than from indigenous goats? Doctoral Thesis Georg-August University of Goettingen, Germany Bass Werner, M. 2004: Personal comunication, 12.10.2004 Berrio, V.A.; V.R. Funquen and L.H. Grajales 1995: Comportamiento productivo de una explotacion de cabras lecheras en la sabana de Bogota. Memorias I Convencion internacional de cabras lecheras y camelidos sudamericanos, Escuela Superior Politecnica de Chimborazo, Riobamba, Ecuador Burrows, J. 1994: Situacion actual y perspectivas de la produccion caprina en Chile. Memorias VII Reunion Nacional de Produccion Caprina, INTA Bariloche, 2 al 4 de noviembre 1994, p 11-12, San Carlos de Bariloche, Argentina Campero, J.M. 1996: Caracterización de índices de producción de genotipos nativos, razas introducidas (Nubia) de caprinos y validación de tecnologías de manejo y procesamiento de productos lácteos. In: Ministerio de Desarrollo Económico, Secretaria Nacional de Agricultura y Ganadería, Instituto Boliviano de Tecnología Agropecuaria IBTA, Informe Anual Gestión 95/96, Programa Ganadería y Forrajes, Cochabamba, Bolivia Cautin, F. 2004: personal comunication, 13.10.2004 Caballero, R. 1994: Sistemas caprinos y ovinos de producción en los valles de Mizque y Aiquile (Cochabamba, Bolivia). En: Iñiguez, L. y E. Tejada (eds.): Producción de rumiantes menores en los valles interandinos de Sudamérica. Memorias de un taller sobre metodología de la investigación. Tarija, Bolivia Chumacero, E. 2004:, personal comunication, 2.11.2004. CID (Centro de Información para el Desarrollo) 1996: Anuario Estadistico del Sector Rural 1995-1996, La Paz, Bolivia

Clark, S. and Sherbon, J. W. 2000. Alphas1-casein, milk composition and coagulation properties of goat milk. Small Ruminant Research. 38. 123-134. CODETAR (Corporación Regional de Desarrollo de Tarija) 1992: Diagnóstico de situación de la ganadería caprina en la Provincia Gran Chaco. Tarija, Bolivia REFERENCES 389

Comité de Coordinación Interinstitucional, 1987: Unidades de Investigación, Extensión y Producción. Facultad de Ciencias Agrícolas y Pecuarias, Universidad Mayor de San Simón, Impresión CERES, Cochabamba, Bolivia Condori, D. 2000: Desarrollo corporal de caprinos Anglo Nubia, Criollo, F1(Anglo Nubia x Criolla) y F1(Angora x Criolla) bajo un sistema de manejo semi extensivo en el valle de Cochabamba. Graduate Thesis, Faculty of Agronomy, Cochabamba, Bolivia Devendra, C. 1972. The composition of milk of British Alpine and Anglo-Nubian goats imported into Trinidad. Journal of Dairy Research. 39. 381-385. Devendra, C. and Chenost, M. 1973: Goats of the West Indies. Z. Tierzuechtg. Zuechtungsbiol. 90: 83-93 Dimassi, O., Neidhart, S., Carle, R., Mertz, L., Migliore, G., Mané-Bielfeldt, A., Valle Zárate, A. 2005. Cheese production potential of milk of Dahlem Cashmere goats from a rheological point of view. Small Ruminant Research. 57, 31-36. Eisele, J. 2001: Die Bedeutung der Anglo Nubier-Ziege fuer kleinbaeuerliche Betriebe in Laendern der Tropen und Subtropen. Graduate Thesis, University of Hohenheim, Germany Ergueta, R. 2004: personal comunication, 14.10.2004 FAO, 2003: http://faostat.fao.org Foyanini, C. 2005: personal comunication, 10.2.2005 Galal, E.S.E. 1995: Goat production systems in Egypt. In: A. El Aich, S. Landau, A. Bourbouze, R. Rubino and P. Morand-Fehr (eds.): Goat production systems in the mediterranean. p161-165, EAAP Publication No. 71, Wageningen, The Netherlands Gall, C. 1996: Goat Breeds of the World. CTA, Margraf Verlag, Weikersheim, Germany, pp 132 Gall, C. 2001: Ziegenzucht. Verlag Eugen Ulmer, Stuttgart, Germany, pp 74 Gallo, S. C.and C.I. Wainnright 1995: Algunas caracteristicas fenotipicas de rebaños de cabras criollas de la IX y X regiones de Chile y peso al nacimiento de sus crias. Avances en Ciencias Veterinarias, Vol. 10 No. 3, 27-31 Garcia, B.O.; E.B. Garcia; J. Bravo and B. Kennedy 1996: Analisis de un experimento de cruzamiento usando caprinos criollos e importados. IV. Fertilidad y prolificidad. Rev. Fa. Agron. (LUZ). 1996, 13: 443-455 Gomez, O.I. 2003: Mejoramiento de la poblacion caprina criolla de Zapotillo mediante la selección y el mestizaje con machos de la raza Anglo Nubian. In: VI Congreso iberoamericano de razas criollas y autoctonas y IV Simposio iberoamericano sobre conservacion y utilizacion de recursos zoogeneticos. Resumenes de ponencias y comunicaciones. 1 a 4 de diciembre de 2003, Recife, Brazil Heinkel, J. 1986: Untersuchungen ueber Milchleistung und Koerpergewicht an Kreuzungen aus Milchrassen und Criollo Ziegen in Venezuela. Graduate thesis, University of Hohenheim, Stuttgart, Germany Hinojosa, O. 2003: personal comunication, 9.7.2003 Hinojosa, R. 2005: personal comunication 3.2.2005 390 REFERENCES

IICA, 1989: Estudio sobre la comercializacion de caprinos en el area del Proyecto Cotagaita, San Juan del Oro. Proyecto de Desarrollo Agricola Cotagaita, Potosi, Bolivia Iñiguez, L. 1989: Diagnóstico de los sistemas de producción asociados con pequeños rumiantes (ovejas y cabras) en los valles del distrito sur de Cochabamba. Programa de Desarrollo Alternativo de Cochabamba, Bolivia Lanari, M.R.; H. Taddeo, E. Domingo, M. Perez Centeno and L. Gallo 2003: Phenotypic differentiation of exterior traits in local Criollo Goat Population in Patagonia (Argentina). Arch. Tierz., Dummerstorf 46 (2003) 4, 347-356 Landau, S.; A. Perevolotsky, Y. Carasso and D. Rattner 1995: Goat husbandry and production systems in Israel. In: A. El Aich, S. Landau, A. Bourbouze, R. Rubino and P. Morand-Fehr (eds.): Goat production systems in the mediterranean. p136-159, EAAP Publication No. 71, Wageningen, The Netherlands Mason, I.L. 1981: Breeds. In: Gall (ed.), Goat Production. Academic Press, London, UK Mason, I.L. 2002: World Dictionary of Livestock Breeds, Types and Varieties. Rev. by Valerie Porter. 5 ed. Wallingford, CABI Publ. 2002-XVIII, 3805 Mendoza, B. 2004: personal communication, 27.11.2004 Montaldo, H.; A. Juarez, J.M. Berruecos, F. Sanchez 1995: Performance of local goats and their backcrosses with several breeds in Mexico. Small Ruminant Research 16 (1995) 97-105 Montaldo, V.H.H. and Meza, H.C.A. 2000: Goat genetic resources in Mexico: Bio- economical efficiency of local and specialised genotypes. Wool Technology and Sheep Breeding. 49: 345-350 Mosua, F. 1989: Como mejorar la producción del ganado caprino? Manual del productor campesino. SONU, Cochabamba, Bolivia Mowlem, A. 1992: The role of artificial insemination for improving goat production in difficult environments such as the tropics. Recent Advances in Goat Production 1117- 1125 Mueller, J. 1994: Los recursos geneticos caprinos locales y exoticos y su potencial. In: Iniguez, L. and Tejada, E. (eds.): Produccion de Rumiantes Menores en los Valles Interandinos de Sudamerica. Memorias de un Taller sobre Metodologias de la Investigacion. Tarija, Bolivia, 16-21 de Agosto de 1993. La Paz, Bolivia. p 74-82 Narvaez, R.F. and E.C. Hernandez 1995: Comportamiento productivo de los caprinos de la zona centro Loja (Cantones de Gonzanama y Paltas). Memorias I Convencion internacional de cabras lecheras y camelidos sudamericanos, Escuela Superior Politecnica de Chimborazo, Riobamba, Ecuador Pari, A. 1998: Producción de leche y crecimiento de caprinos Anglo Nubia y Criolla bajo un sistema de manejo semi-extensivo en el valle de Cochabamba. Graduate Thesis, Faculty of Agronomy, Cochabamba, Bolivia Pariacote, F.A. and Ruiz, L. 2004: Diversity and performance of the goat genetic resources in Venezuela. 8th International Conference on Goats, 4-9 July 2004, Pretoria, South Africa REFERENCES 391

Pariacote, F.A. 1992: Productivity of goat native, alpine and nubian breeds and their crosses in Venezuela. Arch. Zootec. 41 (extra): 555-562 Peacock, C. 1996: Improving Goat Production in the Tropics. A Manual for Development Workers. Oxfam / Farm Africa, Oxford, UK, p 278 Pinaya, J. 2002: Cursos de Capacitación. Manejo de ganado caprino. DRIPAD-PMA, Tarija, Bolivia PMA-DRIPAD, 2003: Unpublished project report, Chuquisaca, Bolivia Porter, V. 1996: Goats of the World. Farming Press, Ipswich, UK Primov, G. 1984: Goat production within the farming system of smallholders of Northern Bahia, Brasil. Department of Rural Sociology, University of Missouri, Columbia, USA Quartermain, A.R. 1983: Evaluation and Utilization of Goat Breeds. In: K. Maijala (ed.): World Animal Science B8 Disciplinary Approach. Genetic Resources of Pig, Sheep and Goat. Elsevier, Amsterdam Reinhardt, R.M. and Hall, A. 1978: Nubian History. America and Great Britain. Second Edition Revised, Hall Press, San Bernardino, California, USA Rewe,T.O.; P.B. Ogore and A.K. Kahi 2002: Integrated goat projects in Kenya: Impact on genetic improvement. 7th World Congress on Genetics Applied to Livestock Production, August 19-23, Session 25, Communication No. 25-19, Montpelier, France Ribas, M.; Gutierrez, M and Hernandez, F. 2000: La cabra criolla cubana: El comportamiento reproductivo de un genotipo inexplorado. In: Memoria. V Congreso Iberoamericano de razas autoctonas y criollas, I Taller Internacional de Ovinos de Pelo. I Taller Internacional de Cerdos de Origen Iberico. 28 de noviembre al 1 de diciembre del 2000, La Habana, Cuba Ribeiro, M.N. 2005: personal comunication 10.6.2005 Sanabria, O.; J. Chacon and C. Cossio 1992: Proyecto de desarrollo caprino en las provincias Mizque y Campero. Estudio de factibilidad. Programa de Dearrollo Alternativo Regional. Cochabamba, Bolivia SENA, 1991: Criador de caprinos. Ministerio de Trabajo y Seguridad Social, Bogota, Colombia Stemmer, A. 1994: Internal report 24.6.1994, Faculty of Agronomy, Cochabamba, Bolivia Stemmer, A. 1996: Internal report 4.6.1996, Faculty of Agronomy, Cochabamba, Bolivia Stemmer, A. 2001: Internal report 6.11.2001, Faculty of Agronomy, Cochabamba, Bolivia Stemmer, A. 2003: Recomendaciones para la sostenibilidad de los módulos caprinos DRIPAD-PMA. Informe final de consultoria. Cochabamba, Bolivia Sripongpun, S. 1991: The status of goat production in Thailand. In: Panandam, J.M.; Sivaraj, S.; Mukherjee, T.K.; Horst, P. (eds.): Goat Husbandry and Breeding in the Tropics. Compiled papers presented in an international seminar carried out by German Foundation for International Development (DSE) at the Institute for Advanced Studies, University of Malaya, Kuala Lumpur. Zentralstelle fuer Ernaehrung und Landwirtschaft, Feldafing, Germany 392 REFERENCES

Vaca, J.L. 2004: Personal comunication, 7.10.2004 Velez, N.M. and Callacna, C.M.A. 1984: Preproductive Behavior of Anglo Nubian Goats in an Arid Environment in Peru. CEIBA, 25 (2): 101-112 Videz, R.J.N. and C. Cardona 1992: Comportamiento reproductivo en cabras fecundadas por inseminacion artificial. Memorias X Reunion nacional de ABOPA, october 1990, La Paz, Bolivia, p 26-29 Yagil, R. 1991: Sustainable goat production in Bolivia. Proposal submitted to AID/CDP- MASHAV Zaibet, L.; P.S. Dharmapala; H. Boughanmi; O. Mahgoub; A. Al-Marshudi 2004: Social changes, economic performance and development: the case of goat production in Oman. Small Ruminant Research 54, 131-140 CONTACT ADDRESSES 393

5 CONTACT ADDRESSES Bass Werner, M. Cautin, F. Director of DRIPAD Tarija, Administrator of SONU (Sociedad Avenida Defensores del Chaco s/n, zona Nueva) aeropuerto, Tarija, Bolivia. Calle Mama Ocllu No. S-0905, Casilla Tel.: ++591 46645837 4231, Cochabamba, Bolivia. email: [email protected] Tel.: ++591 44250562 email: [email protected] Chumacero, E. Ergueta, R. Formerly Head of IBTA Research Station Formerly technician of CEDEAGRO, Chinoli, Potosi; currently lecturer at the Mizque, Cochabamba, currently Director Faculty of Agronomy, of DRIPAD Cochabamba University Tomas Frias, Avenida del Avenida Aroma No. 327, Cochabamba, Maestro s/n, Potosi, Bolivia. Bolivia Tel.: ++591 46225414 Tel.: ++591 44251565 (no email) email: [email protected] Foyanini, C. Hinojosa, R. Owner of Residencial Bolivar and goat National Director of Heifer International farms in Pailon and San Javier, Santa Bolivia, Avenida Moscu s/n entre 5 y 6 Cruz, Bolivia anillo, Santa Cruz, Bolivia Tel.: ++591 3325989 Tel.: ++591 33557235 email: [email protected] email: [email protected] Hinojosa, O. Mendoza, B. Formerly technician of DRIPAD Lecturer at the National University of Cochabamba in Novillero, currently Loja, Ciudadela Universitaria, La Argelia technician of Aldeas SOS, Cochabamba Casilla 795, Loja, Ecuador km 1 a Tiquipaya, zona Linde s/n, Tel: ++593 7574054 Cochabamba, Bolivia Email: [email protected] Tel.: ++591 44421935 email: [email protected] Vaca, J.L. Ribeiro, Maria Norma Lecturer at the Veterinary Faculty of Departamento de Zootecnia University Gabriel Rene Moreno, Universidade Federal Rural de Avenida Centenario, Campus Pernambuco Universitario, Casilla 702, Santa Cruz, Av. D. Manoel de Medeiros, SN, Dois Bolivia. Irmaos, CEP: 51171-900 Recife, Brazil Tel.: ++591 3537676 e-mail: [email protected] email: [email protected] 394

GENE FLOW IN ANIMAL GENETIC RESOURCES. A STUDY ON STATUS, IMPACT AND TRENDS

Editors: Anne Valle Zárate, Katinka Musavaya and Cornelia Schäfer

Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, Germany

ANNEX 9.8

UNIVERSITY OF HOHENHEIM Institute of Animal Production in the Tropics and Subtropics

Boran and Tuli cattle breeds – Origin, worldwide transfer, utilisation and the issue of access and benefit sharing

S. Homann, J.H. Maritz, C.G. Hülsebusch, K. Meyn, A. Valle Zárate

VERLAG ULRICH E. GRAUER ‚ Beuren ‚ Stuttgart ‚ 2005 Sabine Homann, Jacobus H. Maritz, Christian G. Hülsebusch, Klaus Meyn, Anne Valle Zárate: Boran and Tuli cattle breeds – Origin, worldwide transfer, utilisation and the issue of access and benefit sharing

VERLAG GRAUER, Beuren, Stuttgart, 2005. ISBN 3-86186-498-3

© 2005 Institut für Tierproduktion in den Tropen und Subtropen Universität Hohenheim (480a), 70593 Stuttgart, Deutschland E-mail: [email protected]

All rights reserved.

Printed in Germany. Druck: F. u. T. Müllerbader GmbH Forststr. 18, 70794 Filderstadt, Deutschland

VERLAG ULRICH E. GRAUER Linsenhofer Str. 44, 72660 Beuren, Germany Tel. +49 (0)7025 842140, Fax +49 (0)7025 842499 Internet: http://www.grauer.de/, E-Mail: [email protected] This case study is an independent part of the gene flow study implemented by the Institute of Animal Production in the Tropics and Subtropics of the University of Hohenheim. The

Federal Ministry for Economic Cooperation and Development (BMZ) and the German Technical Cooperation (GTZ) acted as commissioner and project executing agency. The Food and Agriculture Organisation (FAO) acted as a support agency. An advisory panel composed of international scientists, representatives of donor and development agencies, the private sector and NGOs closely accompanied the study.

ACKNOWLEDGEMENTS A variety of people from different countries and institutions contributed to accomplish this case study by providing helpful information and data on the gene flow of the Boran and Tuli cattle breeds. These are Tezera Getahun and Ahmed Abdi from the Pastoral Forum Ethiopia, Workneh Ayalew from the International Livestock Research Institute in Ethiopia, and Morné de la Ray from EMBRYO PLUS in South Africa. Our thanks go also to Geoff Maynard from the Mount Eugene Belmont Reds & 5 Star Senepols in Australia, Tafesse Mesfin, and Siboniso Moyo from the Department of Livestock Production and Andrew Mushita from Commutech, both in Zimbabwe. Finally, we wish to thank John Vercoe from Australia, Jacob Wanyama from the ITDG-EA in Kenya, and Kerstin Zander from the Center for Development Research in Bonn, Germany. We gratefully acknowledge contributions from all before mentioned institutions and persons to this study.

TABLE OF CONTENTS 401

TABLE OF CONTENTS List of tables and figures 402 Terms and abbreviations 403 Executive summary 405 1 Introduction 407 2 Overview on Boran and Tuli cattle breeds 408 2.1 Origin 408 2.1.1 Boran 408 2.1.2 Tuli 409 2.2 Description and performance 410 2.2.1 Unimproved Boran 410 2.2.2 Improved Boran 411 2.2.3 Tuli 413 3 Current status and breeding organisation 414 3.1 Utilisation in Africa 414 3.1.1 Population status in the area of origin 414 3.1.2 Potential within Africa 416 3.2 Formation of cattle breeding organisations and programmes 418 3.3 Governmental impact on breeding and trade 422 4 Worldwide transfers of Boran and Tuli live cattle, semen and embryos 427 4.1 Transfers within Africa 427 4.2 Import into Australia and America 429 4.2.1 Australia 429 4.2.2 America 432 4.3 Diagrammes of Boran and Tuli transfers 435 5 Mode of transfer 440 5.1 Traditional and modern transfer mechanisms 440 5.2 Issue of access and benefit sharing agreements 442 6 Conclusions 444 7 References 446 7.1 Personal communications 446 7.2 Internet 446 7.3 Literature 448 8 Contact addresses 456

402 LIST OF TABLES AND FIGURES

LIST OF TABLES AND FIGURES Table 1: Performance of Ethiopian Boran, Improved Boran and Tuli cattle 412 Table 2: Overview on Boran and Tuli breeding organisations worldwide 422 Table 3: Transfers of Improved Boran and Tuli cattle breeding material from eastern and southern Africa 436

Figure 1: Unimproved (left) and improved (right) Boran cattle 412 Figure 2: Worldwide transfers of Unimproved and Improved Boran cattle breeding material from Eastern and Southern Africa 435 Figure 3: Worldwide transfers of Tuli cattle breeding material from Southern Africa 435

TERMS AND ABBREVIATIONS 403

TERMS AND ABBREVIATIONS ACIAR Australian Centre for International Agricultural Research AML African Model Law BCBS Boran Cattle Breeders Society CAIS Kenyan Central Artificial Insemination Station CBD Convention on Biological Diversity CBPP Contagious Bovine Pleuro-Pneumonia CSIRO Commonwealth Scientific and Industrial Research Organisation EARO Ethiopian Agricultural Research Organisation FAO Food and Agriculture Organization of the United Nations FMD Foot and Mouth Disease IETS International Embryo Transfer Society ILRI International Livestock Research Institute KARI Kenya Agricultural Research Institute KETRI Kenya Trypanosomias Research Institute LMA Livestock Marketing Authority LMDE Livestock and Meat Development Enterprise MARC US Meat Animal Research Centre MJ Megajoule NATA North American Tuli Association NGO Non-Governmental Organisation OADB Oromiya Agricultural Development Bureau OAU Organisation of African Unity SADC Southern African Development Community SAABCO South African Australian Breeders Company SA TCBS South African Tuli Cattle Breeders Society

EXECUTIVE SUMMARY 405

EXECUTIVE SUMMARY African indigenous cattle breeds, particularly the Boran and the Tuli, have received increasing interest in the past as a source of genetic diversity with potential to improve cattle production in sub-/tropical environments worldwide. In this context, controversy arises about conservation-through-utilisation strategies and access and benefit sharing. We have studied Boran and Tuli cattle and their past development, their actual utilisation, and their future potential. Information on breed history, current status and breeding organisation, and on worldwide transfers of live animals, semen, and embryos of the breeds was gathered from the scientific literature, project reports and internet sources. Semi- structured interviews were held with key people involved in the breeding of, research into, and utilisation and conservation of the Boran and Tuli. The Boran (Bos indicus) is a major cattle breed in eastern Africa, originating in the Borana rangelands in southern Ethiopia. The Tuli cattle (Bos taurus) decend from a small nucleus herd of yellow Ngwato cattle in Zimbabwe. Both breeds have evolved under harsh arid and semi-arid range conditions, and out of the local herders’ indigenous breeding and selection strategies. In Kenya, rapid breeding of the Boran cattle was started by British settlers, leading to the formation of the Improved Boran. In Zimbabwe, the Tuli were further developed at governmental breeding and research stations from the early 20th century. Breeding organisations were founded in the regions of origin, and research and breeding programmes were set up to improve beef production while maintaining adaptability to environmental constraints. Boran cattle showed high fertility and production in low input environments, while Tuli cattle also showed high fertility and excellent beef quality, but produced comparatively better in high input environments. Because of their adaptability and productivity in tropical conditions, Boran and Tuli cattle attracted the interest of livestock scientists and the international beef industry. In 1988, in order to add breeding options to composite crossbreed populations for industrial beef production, Australian researchers - collaborating with a consortium of Australian beef producers – imported the first Boran and Tuli embryos from Zambia and Zimbabwe. In 1991, Boran and Tuli embryos were exported from Australia to research stations in Nebraska and Texas, USA, where the largest germplasm programmes on beef cattle in the world were undertaken. The breeds also found their way into the Australian, American, and South American beef industries through various other channels. The majority of documented transfers of Boran and Tuli genetic material represent commercial transfers between government research stations and private enterprises and among international business partners. The Tuli appears to have been better accepted among beef producers in Australia and America than the Boran. However, comprehensive data about the current population and the contribution of both breeds in crossbreeding schemes or in the formation of composite breeds are scarce. From the few available figures, the population of these breeds in these countries is small. Given the overall size of the beef industry, with 94.9 million head of beef cattle in the USA and 26.4 million head in Australia in 2004, the likely contribution of Boran and Tuli cattle to the sector appears insignificant. However, the study shows that a demand exists for genetic material with adaptive traits and special traits of beef quality. 406 EXECUTIVE SUMMARY

In Africa, Boran and Tuli cattle were used systematically to a considerable extent for commercial ranch development, but not for local or regional livestock improvement schemes. Today, Boran and Tuli cattle are said to face genetic dilution in their areas of origin, although they are still used – particularly the Improved Boran in Kenya - for commercial beef production. Some Non Governmental Organisations argue that the Boran and Tuli were exploited by networks of institutions, business companies and individuals, without sufficient compensation to the original breeders. They have raised a controversial debate on access and benefit sharing agreements, which is dominated by three main issues: Whether or not a prior informed consent existed before the initial transfers from Africa into Australia; whether or not the call for additional compensatory payments to the original breeders is justified; and how big the impact of Boran and Tuli genetic material on the Australian and North American beef sector actually was. The first two issues are ethical rather than scientific, and have to be politically resolved. The actual contribution of the Boran and Tuli to upgrading the Australian and American beef sector appears to be negligible. It did not nearly meet the high expectations which triggered the operation, and the actual use of Boran and Tuli cattle in the Australian and American beef sector is limited to singular cases. Prior informed consent should be more carefully looked for in future transactions to avoid a posteriori claims.

INTRODUCTION 407

1 INTRODUCTION African indigenous cattle breeds are a source of genetic diversity that is being increasingly recognised for its potential for cattle production in tropical environments worldwide. In their native environment these breeds experience genetic loss due to environmental degradation, recurrent droughts, indiscriminate crossbreeding, replacement by exotic breeds, and the absence or disappearance of breed development programmes. In some industrialised countries, there appears to be a potential demand for some of these breeds. However, although their greater dissemination has become possible with improved disease control and the use of embryo transfer and artificial insemination in their countries of origin, their use has still remained limited. The present case study attempts to describe the worldwide transfer of genetic material of two African cattle breeds, the Boran from East Africa and the Tuli from Southern Africa. The so-called Improved Boran is the result of intensive ongoing breeding initiated by British settlers in Kenya from the indigenous Boran cattle stock of southern Ethiopia, southeastern Somalia and northern Kenya. The Tuli evolved from indigenous Sanga cattle from Zimbabwe and eastern Botswana, and was maintained and improved on state breeding and research stations in Zimbabwe. Available information was gathered from the scientific literature and from the internet. Networks involved in using and conserving cattle genetic resources were identified and key persons were contacted for information. The aim of this study is to give an overview of the origin and movements of the two breeds. Their worldwide distribution, performance and further development in other countries are reviewed. The formation of Boran and Tuli cattle breeders’ societies and their national and international networks are reviewed, as well as governmental breeding and trade policies and their impact on the further development of the breeds. Data on imported and exported germplasm as well as the current livestock populations are quantified for each country where possible. Issues concerning the modes of transfer of genetic material and access and benefit sharing for livestock genetic resources are addressed. 408 OVERVIEW ON BORAN AND TULI CATTLE BREEDS

2 OVERVIEW ON BORAN AND TULI CATTLE BREEDS

2.1 Origin

2.1.1 Boran Boran1 cattle are of the large East African short-horned zebu type with a large thoracic hump and pronounced dewlap, and are commonly classified as Bos indicus. Based on molecular genetic analysis, Frisch et al. (1997) present evidence of East African as ‘taurindicus’, i.e. having both Bos taurus and Bos indicus in their ancestry. Hanotte et al. (2000) found only zebuide alleles in Boran cattle, indicating that the indigenous taurine African Y chromosome has been almost eliminated from the Ethiopian region. According to Gibson (2005, pers. comm.) molecular genetic analysis indicates about 20-30% taurine genes in the Boran, with higher proportions in the improved than in the unimproved (cf. below for the distinction). This is likely due to crossing with European Bos taurus cattle when developing the so-called “Improved Boran”. The thoracic-humped zebu cattle descend from the secondary cattle domestication in the ‘fertile crescent’ south of Mesopotamia at about 5,000 BC (Payne and Wilson, 1999). Archaeological records indicate that they are the most recent types of cattle introduced into Africa from western Asia. Recent molecular genetic and archaeological evidence (Marshall, 2000; Hanotte et al., 2002) suggest that zebu cattle were introduced into Africa through East Africa rather than through the land connection between Egypt and the Near East. After the Arab invasion (670 AD) they were imported into eastern Africa in large numbers and were spread by Indian and Arabian merchants across the Red Sea to drier agro-ecological regions in the Horn of Africa. Over time they displaced the Sanga type cattle as far south as the Zambezi river owing to their greater resistance to Rinderpest and their higher milk yield (Mason, 1984; Loftus and Cunningham, 2000). The Ethiopian Boran cattle originate from the pasture planes at Liban and Dirre, the heartland of the Borana rangelands in southern Ethiopia2. Molecular characterisation suggests that all Boran types originate from this same ancestral stock (Rege et al., 2001). Boran cattle are kept by the Borana pastoralists in a dryland environment with comparatively productive pastures (Pratt and Gwynne, 1977; Coppock, 1994; Oba, 1998). In the past the Boran cattle spread from Ethiopia to south-western Somalia, where they were called Somali Boran or Äwai. The Somali Boran are now mainly found in south eastern Ethiopia, Jubaland in the southern part of Alto Giuba along the Ethiopia - Somalia - Kenya border and also in the Wajir region of Kenya, where they are kept by the Arti and Mohammed Zubier Somali tribes (Felius, 1995; Rege et al., 2001). The Borana and Somali pastoralists migrated with their cattle south-eastward into northern Kenya where the Kenyan Boran then evolved. In the 15th century, Boran cattle moved further south until they arrived west of the river Tana in Kenya. There, the Orma pastoralists developed the more trypanotolerant variety, the Orma Boran, also known as Tanaland Boran (Dolan et al.,

1 The name ‘Boran’ stems from the original habitat, the Borana rangelands in Ethiopia. 2 Earlier literature uses the term Sidamo for the Borana rangelands.

OVERVIEW AND BORAN AND TULI CATTLE BREEDS 409

1994; Rege et al., 2001). According to Felius (1995) the Boran influences even reached Massai territory in southern Kenya. However, the southward spread appears to have come to a halt on the borderline between Ukambani and Masailand in Kenya, where a number of European ranchers took to breeding Boran cattle. Homann (2004) found that the Borana pastoralists in southern Ethiopia differentiate two types, which they call the “Qorti” and the “Ayuna”. The large-framed Qorti is considered as the true Ethiopian Boran. It is appreciated for its high productivity but it is poorly adapted to scarce grazing. The smaller Ayuna is less productive but better adapted to forage scarcity. According to the pastoralists, the Qorti originated in the eastern plains of the Borana rangelands, while the Ayuna type evolved under the gradual inflow of animals from the northern Ethiopian highlands. According to Mesfin (2004, pers. comm.) the representative type of Boran cattle is currently found in the south-eastern Borana - Somali cross-border region, while in many other areas the breed characteristics have been diluted. The so-called “Improved Boran” was developed by British ranchers in central Kenya from pastoral Boran cattle of southern Ethiopia, south-eastern Somalia and northern Kenya. Lack of success in breeding European cattle in Machakos, Laikipia, and the Rift Valley led the ranchers to buy Boran cattle from pastoralists in the 1920s via Somali traders. Through selection and crossbreeding with European beef breeds used by the white farmers at the time (mainly Hereford, Shorthorn, and Simmental) the Boran was then improved to an excellent beef breed for this area. It is during this process that the Improved Boran must have acquired its higher portion of Bos taurus genes. Export of Improved Boran breeding stock to Zambia and other African countries started in 1947, and in 1973 the Improved Boran were quite popular among white farmers in Zambia. Since 1991 the Improved Boran found its way to Australia, North- and South-America for research and commercial purposes (Felius, 1995; Rege et al., 2001).

2.1.2 Tuli The Tuli3 cattle have evolved from the southern African Ngwato cattle, a large heavy boned Sanga cattle type with long horns. Sanga type cattle are commonly classified as genetic intermediates between Bos indicus and Bos taurus breeds. Frisch et al. (1997) support classifying Sanga as being partly Bos taurus, with the Tuli in particular having taurine but no Bos indicus ancestors. This was confirmed by Hanotte et al. (2000) who found taurine dominance among Sanga breeds in the southern African region. Sanga cattle are commonly thought to originate from hump-less long-horned cattle, but with zebu influences in the Horn of Africa between 5,000-1,500 BC (Oliver, 1983). The Sanga cattle moved with the human migrations to new areas in southern Africa. The Bantu crossed the Zambesi around 700 AD and remains of Sanga cattle were found in Zimbabwe dating back to 900 AD (Epstein, 1971). The cattle were then moved from Botswana in the west to Mozambique and Zululand in the east. From west of Zimbabwe they were taken to South Africa prior to the 15th century. From this cattle type European settlers developed the

3 The name tuli stems from the Ndebele word ‘utulili’ meaning ‘dust’, and refers to the arid native environment of the Tuli cattle. 410 OVERVIEW ON BORAN AND TULI CATTLE BREEDS

Africander breed. Later the Bonsmara breed resulted from crosses with Hereford and other shorthorn cattle (Mason 1984). Hetzel (1988) reports that Tuli cattle originate from the Tswana cattle indigenous to the vast arid sweet-veld region in the central and northern part of southern Africa and southern part of central Africa. He considers the Tuli as an improved Tswana breed. According to Felius (1995), the Bamangwato livestock farmers, who originally settled in south-eastern Botswana and then spread across the Transvaal and southern Zimbabwe before the advent of the Matabele livestock farmers, had a distinct type of cattle. These Bamangwato cattle, a Sanga strain called Ngwato, formed one strain of what is named Tswana, but were almost extinct following the Rinderpest outbreaks in 1896. In 1942 a land development officer, Mr L.V.A. Harvey, observed a distinct type of yellow Sanga cattle among the mixed local Ngwato cattle herds. These yellow Ngwato seemed to be better adapted to local conditions and superior beef animals. Harvey persuaded the Zimbabwe government to start buying a basic breeding herd of the last remaining Tswana type Ngwato cattle from the local livestock owners in the Tuli area, in south western Zimbabwe (TCBS, 2004). In 1945 a state breeding station was established on about 1,200 hectares in Matabeleland with a founding Tuli herd of twenty cows and one bull, because these already possessed desirable characteristics such as high fertility and good ease of milking. Further selection efforts focused mainly on heritable traits concerning meat production and conformation, and on naturally polled animals of a golden brown colour. The result was the rapid development of what was described by Goodwin (1976) as a highly fertile purebred indigenous breed with good beef and milk qualities. In 1950 the government decided to extend the research station to about 8,000 hectares and by 1956 the Tuli numbers had risen to 1,000 (Harvey, 1987). The original breeding policy was designed to supply improved animals to the smallholder sector in the low rainfall areas. However, over the years commercial ranchers were increasingly attracted by the Tuli and became involved in its breeding (Harvey, 1987). Due to security problems from the liberation war in Zimbabwe, the Tuli herd was translocated to Matopos Research Station in south western Zimbabwe in 1979, where breeding continues today. The breed is now mainly used by commercial farmers. The success of the improved Tuli breed has led to the export to neighbouring African countries as well as to Australia and northern and southern America (Mushita, 2005, pers. comm.).

2.2 Description and performance

2.2.1 Unimproved Boran The Boran is the largest of the East African zebu breeds. Mature body weights of unimproved Boran cattle range from 300 to 650 kg in males and from 200 to 360 kg in females (Table 1; on-staion and field data). The typical Ethiopian Boran cattle have a deep chest, long legs, a large and wide body frame, and hard hooves. They are used by pastoralists for milk and meat. The larger Somali Boran is short legged and mainly used for milk. The Orma Boran is more resistant to trypanosomiasis and is the smallest type among the Boran cattle. The coat colour of Boran cattle is usually light grey, white or fawn, reddish or brown, with characteristic dark grey forequarters including the thoracic hump and the pendulous dewlap. The hump is well developed and hangs over to one side in

OVERVIEW AND BORAN AND TULI CATTLE BREEDS 411 males. The udder is well developed in females (Felius, 1995; Reda, 2001; Rege et al., 2001). Survival and fertility were the most important traits for local cattle raisers and stringent natural selection ensured hardy animals. Consequently, milk yield of individual cows was low and large herds were required to ensure sufficient milk supply (Rege et al., 2001). Physiological adaptations to harsh environments are common for zebu breeds. They are known to regulate body temperature more efficiently than taurine cattle and therefore have lower water requirements. They also show more efficient ruminal digestion and protein synthesis along with lower metabolic faecal nitrogen excretion than most temperate breeds. Furthermore, their hard hooves and light bones enable them to endure long migrations (Hunter and Siebert, 1985). The particular characteristics of Boran cattle - with the Ethiopia, Somalia, Kenya and Orma Boran types sharing similar traits - are summarised in the Domestic Animal Genetic Resources Information System (DAGRIS) database of the International Livestock Research Institute (ILRI) as follows: Excellent walking ability, drought resistance, pronounced herd instinct, excellent mothering ability, docile but aggressive with the calf at foot, high disease resistance and tolerance, good heat tolerance, good adaptation to a variety of climates, longevity, high fertility, and pronounced sex dimorphism with small females and large males (DAGRIS, 2005). However, both the unimproved Boran - but particularly the Improved Boran - are generally found to be more succeptible to diseases, ticks, and other environmental stresses than other, smaller East African zebu breeds. Similar to the humpless Longhorns adapting to the tsetse bush by forming the small N’Dama, the zebu became smaller in size when exposed to humidity and disease, the smallest indigenous cattle breeds of Eastern Africa being found in the Mbulu Highlands of Tanzania and the Taita Hills in Kenya. These small East African zebus have lower yields than the unimproved and the Improved Boran (Meyn, 1967; Meyn, 1970).

2.2.2 Improved Boran Improved Boran cattle are primarily kept for beef production and are described as an excellent tropical beef breed with good meat quality (Kenya Beef Industry Development Project, 1974). The Boran Cattle Breeders Society of Kenya describes them as being well adapted to arid environments and showing good productivity under environmental constraints (BCBS, 2005). Mature body weights range from 550 to 850 kg in males and from 304 to 414 kg in females (Table 1). Although morphologically similar to the other Boran types, they typically have well-developed hindquarters (Rege et al., 2001) and well fleshed thighs, especially in the upper parts (Felius, 1995) (cf. figure 1). They can maintain these economically valuable qualities also under unfavourable environments because of their good adaptability to heat, to low forage quality, to low water availability, and to walking long distances. The cows are known to be very fertile with good calving ease and mothering traits (BCBS, 2005). Compared to the Orma Boran, the Improved Boran is less productive under trypanostress (Rowlands, 1995 in Köhler-Rollefson, 2001). Some breeders selected for white breeding herds but red-brown is the predominant coat colour. Rege et al. (2001) found high numbers of alleles diverging from the Hardy Weinberg equilibrium in the Improved Boran, which they attributed to intentionally selecting for beef production. 412 OVERVIEW ON BORAN AND TULI CATTLE BREEDS

Table 1: Performance of Ethiopian Boran, Improved Boran and Tuli cattle

Traits Unimproved Improved Tuli Boran Boran Wither height (cm) Male 111 - 127 133 - 147 Female 113 - 118 115 - 120 Adult Liveweight (kg) Male 300 - 650 550 - 850 Max 800 Female 200 - 360 303 - 414 369 - 468 Age at first calving (month) 34 35 - 47 36 Calving interval (days) 370 - 743 380 - 544 389 - 475 Pregnancy rate (%) 97 90 Calf survival (%) 96 96 Birth weight (kg) 23 - 26 24 - 44 29-32 Weaning weight (kg) 85 - 190 160 - 217 170-184 Age at weaning (days) 252 214 - 26 Age at puberty (days) 389 365 Puberty weight (kg) 310 299 Meat production Carcass weight, 447 days (kg) 310 308 Daily gain, birth-weaning (g) 559 - 636 289 - 408 622 Daily gain, weaning - slaughter (g) 272 Milk production Length of lactation (days) 70 - 319 Total milk yield/off-take (kg) 66 - 513 Daily production (kg) 0.8 – 3.9 0.8

Sources: Bock (1971); Cundiff et al. (1996); DAGRIS (2005); Felius (1995); Meyn (1967); Tonn (1974).

Figure 1: Unimproved (left) and improved (right) Boran cattle

Sources: Homann (2004); Hülsebusch (1992)

OVERVIEW ON BORAN AND TULI CATTLE BREEDS 413

2.2.3 Tuli Tuli cattle are also considered as one of the world’s best-suited breeds for improving beef production under extensive ranching conditions. Tuli have a moderate frame and a colour, which ranges from silver, yellow, light brown to deep red. Few animals are pale grey to black. Although descended from the horned Sanga cattle, about 70-80% of the animals are naturally polled, which indicates taurine influence during breed evolution. The rest have long and occasionally loose horns. The Tuli cattle offer maximum hybrid vigour in crossbreeding programmes with genetically different Bos indicus breeds (TCBS, 2004). Tuli cattle are described as an early maturing beef breed with excellent meat quality and high fertility (TCBS, 2004). Tuli cows reach body weights of 370-470 kg, bulls of up to 800 kg (Table 1). Well-managed steers weighed about 275 kg at 33 months of age and dress out at 60%. Cows are very fertile and many breed beyond 15 years of age (Felius, 1995). Tuli cattle reached calving rates of 70% and calf mortality rates of only 9% under extensive ranching conditions in Zimbabwe. Their ease of calving along with low birth weights, good mothering instincts, disease and parasite resistance, good feed conversion and their docile temperament are appreciated for ‘maximal production under minimal cost’ (Tawonezvi, 1984).

414 CURRENT STATUS AND BREEDING ORGANISATION

3 CURRENT STATUS AND BREEDING ORGANISATION Today most of the unimproved Boran cattle in Africa are kept in subsistence-based systems in southern Ethiopia, northern Kenya, south-eastern Somalia and Tanzania (Ali, 1998). Improved Boran cattle are mainly kept on commercial ranches on the semi-arid to arid plains of the Rift Valley and the Eastern Provinces of Kenya. They have been exported to Tanzania, Uganda, Zambia, Nigeria and Congo and more recently to South Africa. The pure Tuli cattle in Africa today are mainly raised in the commercial sector of Zimbabwe and Botswana, although the population declined over the last years. South Africa has developed a growing Tuli population. Namibia and Gabon are also reported to have Tuli cattle (Felius, 1995; TCBS, 2004). Usually the subsistence sector keeps most of the indigenous cattle, but no evidence exists as to whether these are Tuli or the related Tswana cattle. After the systematic within-breed selection in Africa, both breeds were imported for commercial or experimental purposes into Australia and North America. Australia started the transfer of embryos from Africa for research and commercial investigations. The USA, importing from Canada and Australia, started the largest research programmes on Boran and Tuli cattle. Some cattle were also transferred to Latin America. Although both breeds clearly showed their value for beef production in the tropics, relatively few producers in Australia and North America adopted the Tuli, and were even more reluctant towards the Boran. The Tuli was reported to also exist in Argentina, the Boran also in Mexico and Brazil. Boran and Tuli breeds play a role for both subsistence and commercial livestock production systems and their roles and uses, the breeding strategies, and the potential for improvements therefore differ accordingly. The commercial sector was comparatively better served by data from formal registrations and advanced breeding organisations. Available documents were mainly on the performance of Boran and Tuli stock, however, the commercial impact does not seem to be established yet. Data on the subsistence sector, taking ecological and socio-economic changes into account, were very limited and can only be presented as general trends. The following chapters therefore refer mainly to the commercial sector in the African countries of origin as well as to the further development of the breeds in Australia and in the USA. Data on Boran and Tuli imports into African countries were scarce and not further investigated. No data were found for Latin America. The nature of this study is to address in more detail those countries and cases for which information is available. Thus it does not represent all Boran and Tuli transfers worldwide, but merely indicates the general processes.

3.1 Utilisation in Africa

3.1.1 Population status in the area of origin The unimproved or indigenous Boran cattle are abundant in southern Ethiopia with an estimated 1.8 million head in 1992, while the Improved Boran is not found there

CURRENT STATUS AND BREEDING ORGANISATION 415

(DAGRIS 2005). A National Artificial Insemination Centre (NAIC) distributes Boran semen within the country since 1980 (Getahun and Ahmed, pers. comm., 2005). DAGRIS (2005) differentiates the indigenous Boran cattle population of Kenya between the Orma Boran with 547,000 head, the Improved Boran with 580,570 head, and the Kenya Boran which has no population data. Two Improved Boran bulls where placed at the Central Artificial Insemination Station in Kabete, Kenya, around 1972 to provide semen for Kenyan Boran breeders and for export. However, the demand for semen remained low. In 1990, 90% of the Improved Boran females were estimated to be purebred and nine males were registered for artificial insemination (FAO DAD-IS, 2005). During 1999 and 2000, about 1,300 new Improved Boran cattle were registered in the Kenya Stud Book (BCBS, 2005). In Zambia the total Boran population is estimated to be less than 100,000. Four males were in artificial insemination service and deep frozen semen is available (FAO DAD-IS, 2005). In Tanzania the total Boran population is estimated to be below 5,000 and falling, with about 2,000 females and 500 males. In South Africa, 61 Improved Boran were registered in 1998, including 30 breeding females (FAO DAD-IS, 2005). No population data were found for Somalia, Uganda, Congo and Nigeria. Tuli cattle are found in southern Zimbabwe and neighbouring countries in the hands of commercial farmers. According to Moyo (2005, pers. comm.) the Tuli is the most popular indigenous cattle breed in the commercial sector. In 1992 the total number of Tuli cattle in Zimbabwe was estimated at 10,000 with 2,380 females registered in the herdbook and eight males used for artificial insemination (FAO DAD-IS, 2005). DAGRIS (2005), however, classifies today’s Tuli population in Zimbabwe as endangered. No population census data were available for Botswana. The government owned a Tuli breeding herd of 60 cows in 1996 and 57 animals were used in a breed evaluation study in 2000 (Mpofu, 2002). In 1990, 800 straws of deep frozen semen were in the semen bank and 1,112 straws were used for artificial insemination (FAO DAD-IS, 2005). Currently, South Africa has the biggest Tuli herd. The population has risen from 1,629 registered females and 564 registered males in 1997 and 2,339 registered animals in 1998 (FAO DAD-IS, 2005) to about 4,000 registered animals in 2004 (Pretorius, 2004). In 1997, one male was registered for artificial insemination and 5,477 doses of semen were stored at a commercial artificial insemination station, while 3 sires were used for cryo-conservation (FAO DAD-IS, 2005). No population data were found for Namibia. Purebred Boran and Tuli populations faced genetic dilution due to various factors in their countries of origin. In Ethiopia this was due to uncontrolled introgression and restocking with inferior highland cattle after drought (Haile-Mariam et al., 1998; Mesfin, 2004, pers. comm.). Degradation of the environment further jeopardises the pure Boran (Homann, 2004). In Kenya the main threats to the Improved Boran are said to be uncontrolled crossbreeding with exotic breeds, disappearing breeding structures, neglected traditional livestock systems and shrinking grazing resources (Rege et al., 2001). Irungu (cited by Daily Nation, 1998) states that about two thirds of ranch cattle classified as Boran were actually crossbreds. The status of the Tuli population in Zimbabwe is unclear. It is regarded at risk on commercial ranches following the national land resettlement programmes. The

416 CURRENT STATUS AND BREEDING ORGANISATION

privately owned breeding nucleus of 200 cattle in Matabeleland was reportedly sold when farmers sold their cattle due to compulsory acquisition of their farms (Gande, 2000). In the light of the current situation in Zimbabwe, the South African Tuli Cattle Breeders Society considers exporting Tuli breeding stock from Zimbabwe unlikely (TCBS, 2004). While Pretorius (2004) claims that there are hardly any improved Tuli cattle left, Moyo (2005, pers. comm.) maintains that although farms have changed owners, the Tuli cattle are still there but have not been officially re-registered. According to Mhlanga et al. (1999) cattle restocking after the 1992 drought favoured exotic breeds and they replaced indigenous breeds in the communal areas. In Botswana, the meat grading system seems to favour large framed exotic breeds over the Tuli. Data from the Ramatlabama Bull Study and Artificial Insemination Laboratory (1987-1995) illustrate the promotion of exotic European bulls at the expense of native cattle, with a very low share of Tuli bulls in natural service (0.5- 3.2%) and artificial insemination (0.2-2.4%) (Nsoso and Morake, 1999). Crossbreeding with European breeds appears to have considerably reduced or diluted local cattle breeds (Senyatso and Masilo, 1992). Before independence in 1966, the livestock industry used mainly native cattle breeds and contributed substantially to agricultural production and export earnings. After independence, the government launched a bull subsidy scheme and artificial insemination services, open to all farmers, which led to the above phenomena.

3.1.2 Potential within Africa African indigenous cattle breeds are recognised as essential for people living in marginal areas where no other land use is possible. Indigenous cattle are often kept in mobile or smallholder systems for multiple purposes, different between locations. Recent studies show that indigenous breeds are often more productive than European breeds in such systems, especially in unfavourable environments. This is attributed to their superior adaptation to environmental stress, including resistance to diseases and parasites. For those systems, developing single purpose breeds to produce either beef or milk, is considered inappropriate (Rege, 1998; Rege et al., 2001). The Boran cattle in Ethiopia are predominantly kept in pastoral systems in the arid and semi-arid lowlands. Pastoralism provides a living for about 10-12% of the country’s human population, and pastoralists keep 40% of the national cattle herd and use 60% of the total area. Figures from the 1990s suggest that they also contribute about 90% of the legal live animal exports of the country, and that they supply 20% of the draught animals used in the Ethiopian highlands (Coppock, 1994; Sandford and Habtu, 2000). Further, a considerable share of Boran cattle is reportedly exported to Kenya by unofficial cross border trade (Teka et al., 1999). Several studies suggest that Boran cattle in indigenous pastoral systems can perform better than commercial beef cattle ranching systems under similar ecological conditions. Borana pastoralists were found to achieve a 57% higher productivity than did Kenyan ranches, when using gross energy edible by humans (in MJ per ha and year) as an indicator (Cossins, 1985). However, this might not hold for dairy ranching systems. Pastoral Boran cattle yielded more milk per cow and year than did Massai cattle (219- 251 kg vs. 50-235 kg) and permitted a slightly higher cash output per head and year

CURRENT STATUS AND BREEDING ORGANISATION 417

(20-27 US$ vs. 16-24 US$) (Bekure et al.,1991; de Leeuw, 1995; Behnke and Abel, 1996). Intensive research was therefore launched in the 1980s on the growth potential of Boran cattle (Arnason and Kassa-Mersha, 1987; Sisay and Ebro, 1988; Tegegne et al., 1994; Haile-Mariam and Kassa-Mersha, 1994), but no commercial beef sector has yet developed in Ethiopia. As crossbreds with Boran cows showed improved milk production, they were thought to have the potential to meet the increasing demand for dairy products in the densely populated highland regions of Ethiopia. On station results showed that milk production of Boran purebreds was too low when compared with exotic dairy breeds kept in commercial highland farms. Milk production of Boran crossbreds with Holstein Friesian or Jersey increased threefold in the first generation due to significant additive and heterotic contributions. However, the high level of production could not be maintained in later generations due to recombination losses. To overcome this problem, the late Douglas Hinde of Lolldaiga Hills Estate in Kenya used criss-crossed milking cows of Boran x Red Poll in a dairy ranching system. Two stud herds of Boran and Red Poll were kept and only purebred bulls were used to serve the crossbred cows. Demeke et al. (2004) recommended a breeding strategy of selecting within purebreds and then producing F1 crosses that could be used as dairy cows. Holstein Friesian x Boran crossbred dairy cows were evaluated for their potential as dual-purpose cattle (milk production and traction) for small scale farmers in the Ethiopian highlands. Milk production and reproduction were not significantly affected by work both on station and on farm (Shapiro et al., 1994), and work output of dairy cows was equivalent or above that required by farmers for land cultivation (Zerbini et al., 1998; Alemu, et al., 1998). Dual-purpose crossbred Boran cows were thus considered an attractive innovation in areas where increasing demand for arable land has reduced the size of land available for natural grazing and browse (Demeke, 1999). Sufficient feed quantity and quality was a major factor determining the performance of dual-purpose cows. In Kenya the majority of Boran cattle – the unimproved Boran - are kept by pastoralists. A considerable number of large-scale private and public ranches also keep Improved Boran cattle on arid and semiarid pasture plains of the Rift Valley and the Eastern Province (Rege et al., 2001). The Improved Boran cattle in particular have an established place in the national beef sector. The Kenya Beef Industry Development Project at Lanet, a Kenya Agricultural Research Institute (KARI) station, chose the Boran as the most common breed in the market as a basis for starting a cattle feeding industry. In comparative feedlot trials between 1968 and 1973 the Improved Boran produced high quality beef and after a 16 weeks fattening period showed a dressing percentage of 52%, grade scores of 4.68 and an average daily gain of 1,098g, while the unimproved Boran had a dressing percentage of 51.7%, grade scroes of 4.24 and an average daily gain of 883 g after a 16 weeks fattening period. From this data, the potential to increase the yield of edible carcass from one unimproved head of pastoral cattle was estimated 30-50% during 10 weeks feedlot. It was thus recommended to use this potential to capitalise upon Kenya’s excellent rangelands, combining ‘cheap growth’ on native pastures with compensatory growth during finishing in the feedlot (Creek, 1977).

418 CURRENT STATUS AND BREEDING ORGANISATION

The Orma Boran was found to be superior in Tse-tse infested areas. Research at the Kenya Trypanosomias Research Institute (KETRI) showed that they produce better than other Bos indicus under Tse-tse challenge. They are less likely to suffer from severe anaemia, recover sometimes without treatment, and infection and mortality rates are half that of the Improved Boran. The programme resulted in Orma calves, with improved birth and weaning weights (bulls reach up to 400kg at 4 years of age), being now sold to farmers in other Tse-tse infested areas (Dolan et al., 1994). In Southern Africa’s commercial sector Tuli cattle are mainly used for beef production, whereas in the communal sector milk, draught power and manure play a more important role (Moyo, 2005, pers. comm.). The Tuli improvement programme in Zimbabwe at Matopos Research Station and research in Botswana confirmed its high fertility and the ability to produce well under extremely diverse natural conditions. The Tuli was rated for its docility and mothering qualities (Trail, 1984; Tawonezvi et al., 1988). The University of Zimbabwe (UniZim, 2005) is characterising the Tuli breed for beef production. Hetzel (1988) ascribed the Boran and Tuli cattle a leading role for beef production in Africa after evaluating six African cattle breeds under conditions. His comparison was based on studies covering a wide range of environments in Uganda (Gregory et al., 1985), Zambia (Thorpe et al., 1980; Thorpe et al., 1981), Zimbabwe (Animal Productivity Research Team, 1969, 1971) and Botswana (Trail et al., 1980; Animal Production Research Unit, 1981). Reproduction and growth traits of purebred Boran, Tuli, Mashona, Barotse, Angoni and Afrikander in comparison with the Brahman were assessed in high- and low-input environments. The indigenous purebreds were highly productive, largely due to their high reproductive rates. In high-input environments the Tuli were the most productive breed per unit of cow weight, but performed poorly in the low-input environment. The Boran performed well in several trials in low-input and hot environments. Crossbreds in this study did not perform better than purebreds. Based on these results, within-breed selection of Boran and Tuli cattle was recommended for national breeding programmes in unimproved environments.

3.2 Formation of cattle breeding organisations and programmes Breeding organisations form out of commercial breeding interests and they provide a regulatory framework to evaluate breeds. Breeding strategies for commercial systems are likely to be different to those targeting subsistence-oriented producers. Many African countries have cattle breeding organisations in the commercial sector. Despite the importance of the subsistence sector, no breeding organisations have evolved in this area. Breed development programmes were formerly strongly technical, but under the challenge to meet the local breeders’ needs, they increasingly emphasised organisational and institutional aspects. The formation of breeding organisations and programmes is described for the countries with a major stake in the Boran and Tuli cattle breed in the following section. In Ethiopia no formal breeding organisation to maintain, use and develop the Boran breed has been established. The Borana indigenous institutions, once famous for complex management systems and regulated rangelands and water resources, have not

CURRENT STATUS AND BREEDING ORGANISATION 419 excelled in breeding Boran cattle. The topic was not raised at the last two Gumi Gayo, an eight yearly pan-Borana assembly to review laws and regulations. However, Shongolo (1995), Huqqa (1999) and Reda (2001) argued that skilled cattle breeding is a part of the daily livelihood management, and of customs and culture. According to Getahun and Ahmed (2005, pers. comm.) establishing a regional Oromia Boran Cattle Breed Association is extremely important to solicit technical and financial support. A rising awareness for the need to ensure the sustainable use of Boran cattle is also indicated by the foundation of the Gayo Boran cattle NGO in 2003, initiated by Sora Adi, himself a Borana and advocate of pastoral development. The breeding programmes implemented by the Ethiopian government were for a long time characterised by top down approaches and destined to cater for the demands of the highland population. They did not take the available resources and the production objectives of the Borana pastoralists into account. The first breeding programmes started in 1972, at the government-owned Abernossa ranch4 in the Ethiopian highlands. Boran cows were crossbred with Friesian cattle to produce F1 heifers for dairy production (Haile-Mariam et al., 1998). Another long-term crossbreeding programme using Boran cows for dairy production was initiated in 1974 at the governmental Holetta Research Station, also in the Ethiopian highlands. Since 1978 the Ethiopian Agricultural Research Organisation (EARO)5 together with ILRI studied Holstein Friesian x Boran F1 crossbred cows for their profitability as dual- purpose dairy-draught animals (Alemu et al., 1998; Demeke, 1999). Crossbreeding studies between 1990 and 2000 used common sires supplied by the NAIC through artificial insemination and natural service (Demeke, 2004). Comparatively little investment is being made in breeding programmes in the beef sector. The Did Tuyura governmental breeding ranch was established in 1987 to conserve and improve the Ethiopian Boran cattle in its native environment. It was supposed to distribute superior bulls to the local pastoralists and to provide Boran purebred heifers to Abernossa ranch for crossbreeding with Holstein Friesian. Due to a lack of funds and facilities the ranch operates below its capacity. Within-breed selection has however been maintained and high quality bulls are distributed to pastoralists at a comparatively low price. According to Zander (2005, pers. comm.), the ranch had about 1,400 animals, including 180 bulls and 600 breeding females at the beginning of 2005. About 60 bulls were distributed to the local pastoralists in 2004. The pastoralists were reported to have a strong interest in buying bulls from the ranch to upgrade their stock. The regional Oromiya Agricultural Development Bureau (OADB) and ILRI started a livestock breed survey in 2000 to characterise the

4 The Abernossa cattle breeding and multiplication center (about 2,500 ha) is one of the two remaining government ranches in Ethiopia. The other ranch is the Did Tuyura Boran indigenous cattle breed improvement and multiplication centre (about 5,000 ha), located in the Borana. rangelands. All other ranches were transferred to private hands in the course of the market liberalisation policy. 5 The national Ethiopian Agricultural Research Organisation (EARO) founded in 1977 as successor of the Institute of Agricultural Research (IAR), has the mandate to coordinate all agricultural research activities in the country.

420 CURRENT STATUS AND BREEDING ORGANISATION domestic animal genetic resources for future improvement (Ayalew and Rowlands, 2004). The situation in Kenya is different. The stakeholders in the beef industry are encouraging cattle producers to take up the Improved Boran. The Boran Cattle Breeders Society (BCBS) and a Boran herd book were established in 1951 (Table 2). The BCBS has introduced standards for registering typical Improved Boran cattle. BCBS still actively promotes the Improved Boran in order to retain the breed’s efficiency and adaptation to harsh environmental conditions. By 2001 the BCBS had 64 paid up members. The BCBS executive committee sets and maintains standards, which are applied by a panel of inspectors (BCBS, 2004). BCBS collaborates with the Kenya Agricultural Research Institute (KARI), the Kenya Central Artificial Insemination Station (CAIS) and the University of Nairobi. ILRI has started a Conservation of Animal Genetic Resources Project, assessing the global genetic resources, identifying unique gene pools and developing tools for economic analysis, in which the Boran is included as important breed (ILRI, 2005). KARI is running a Beef Research Centre in Lanet with the mandate to develop beef production technologies for high and medium potential areas of Kenya. In order to conserve the Boran, the centre advocates within-breed selection, establishing a national Boran stud breeding herd, popularising the breed and involving multiple stakeholders in planning and research. By selling fertile animals to the livestock keepers, they expected to increase the Boran herds' overall fertility, growth and beef production potential. Later, the centre also incorporated bull performance traits into the improvement programme. Livestock keepers were asked to take their choice bulls to a central place for finish feeding over a period of six months, after which they were evaluated for growth, semen quality and concentration (Daily Nation, 1998). They expected to better understand the various parameters that determine growth and reproduction, and so develop simple and accurate methods of selecting breeding bulls (Rege et al., 2001). KETRI embarked on an ambitious breeding and selection programme on the trypanotolerance of the Orma Boran cattle from 1980 to 1996. It was implemented at Galana ranch in the Coast Province of Kenya. Galana borders the Orma pastoral grazing land. Orma Boran cattle were found less susceptible to trypanosomiasis than other breeds, including the Improved Boran. Background information on cattle keeping by Orma pastoralists and on local control measures for trypanosomiasis were collected (Irungu, 2000). The programme aims at involving the community in Tse-tse control and improving beef production while at the same time maintaining the trypanotolerance of the Orma Boran. KETRI plans a multiplication programme for commercialisation of the Orma Boran (Dolan, 1997). Zimbabwe established a Tuli Breeders Society in 1961. The Tuli became part of the Zimbabwe herd book in 1994 and in 2002 seven commercial Tuli breeders were registered. According to Mhlanga et al. (1999) there are currently no national conservation and utilisation programmes for indigenous livestock genotypes. In Botswana the Tuli Breeders Society was established in 1966. The herd book recognised all uni-coloured animals except those being black (FAO DAD-IS, 2005).

CURRENT STATUS AND BREEDING ORGANISATION 421

In the 1990s the Botswana Department of Agricultural Research started a selection programme to conserve cattle breeds, which was necessary for the Tswana rather than for the Tuli cattle, and to improve their productivity, based on two herds with 300 cows each, and so promote establishing cattle breed societies (Senyatso and Masilo, 1992). The South African Tuli herd book and the South African Tuli Cattle Breeders Society (SA TCBS) were established in 1994. In 1999 the SA TCBS gave permission to change the status of the Tuli from a developing breed to an established breed. In 2001 the SA TCBS had 36 registered herds, including one herd in Namibia. The Tuli breeders have started participating in the National Beef Improvement Scheme. About 60% of the registered animals and about 43% of the society members are involved and several farmers gained the prestigious Farmers’ Weekly trophies for best producing cows (Pretorius, 2001; Mpofu, 2002). In 1998 the South African and the Zimbabwean Breeders Societies started to incorporate the data on the Zimbabwe Tuli cattle into those on the South African Tuli cattle to form the South African Breedbook as basis for a joint genetic evaluation programme (Pretorius, 2004). In 2000 the agreement was signed and the BLUP breeding values for 21,647 Tuli cattle became available. The most recent development is a state herd established under the management of the Agricultural Research Council - Animal Improvement Institute at Loskop, South Africa (TCBS, 2005). The privately owned embryo transfer and artificial insemination industry has grown in South Africa. The first and main registered embryo transfer and artificial insemination company in the whole of Africa is Embryo Plus, which started as a family enterprise in 1980. Embryo Plus specialises in bovine embryo transfer and semen collection, transfer technology and trade, as well as trains veterinarians in embryo transfer. 90% of all embryo exports of South African breeds today are through Embryo Plus, with customers in the USA, Brazil, Paraguay, Australia, Bolivia, China, Canada, Argentina and throughout Africa. Embryo Plus operates at five registered quarantine centres in South Africa and has one centre each in Kenya and Zimbabwe. Embryo Plus frequently works in association with numerous embryo centres throughout the world. Ongoing negotiations for protocols are expected to open further alleys for exports of South African genetic material (de la Rey, 2005, pers. comm.; Embryoplus, 2005).

422 CURRENT STATUS AND BREEDING ORGANISATION

Table 2: Overview on Boran and Tuli breeding organisations worldwide Country Society Year of Members Registered cattle foundation (year) (year) Kenya Boran Cattle Breeders 1951 64 1,300 Society (2001) (1999-2000) Zambia Boran Cattle Breeder Not Not Not Society known known known Zimbabwe Tuli Cattle Breeder 1961 7 2,380 fem. (1992) Society (2002) 869 mal. (1992) Botswana Tuli Cattle Breeders 1966 Not known Not Society known South Boran Cattle Breeders 2003 Not 61 Africa Society of SA known (1998) South Tuli Cattle Breeders 1994 36 4,000 Africa Society (2001) (2004) Australia Boran Cattle Breeder 1992 Not 26 Association known (2000) Australia Tuli Cattle Breeder 1990 28 161 Association (2004) (2000) USA, North American Tuli 1998 26 500 cows Canada, Association (2002) (2002) Mexico

3.3 Governmental impact on breeding and trade Livestock breeding legislation regulates animal identification and herd book keeping through performance testing, animal reproduction, genetic material trade, zoo-sanitary and quarantine regulations, and subsidies to livestock production. However, there is not much of a framework for conserving Farm Animal Genetic Resources (AnGR), which includes identifying the status of livestock biodiversity, strategies for conserving and sustainably using AnGR, and economic analysis (FAO, 2004). This chapter reviews current breeding and trade policies on cattle genetic material for the Improved Boran and Tuli in their original countries in eastern Africa as compared to southern Africa. Key policy areas such as governance, land-use policy and infrastructure development, which influence the Boran and Tuli breeds are also considered. Ethiopia and Kenya lack encompassing livestock breeding policies, but promising structures are opening up. Zimbabwe and South Africa, both SADC countries, are further ahead to adjust their policies to international requirements for the management of AnGR. Uganda has made recent efforts in breeding activities and has implemented an animal breeding Act in 2002 (see Nakimbugwe, 2003). African States have an obligation under the WTO TRIPS Agreement to either adopt patents, or a sui generis law, or a combination of both. So, through the Organisation of African Unity (OAU), they have endorsed a model legislation to protect the rights of local communities, farmers and breeders, and to regulate access to biological

CURRENT STATUS AND BREEDING ORGANISATION 423

resources, which is known as the African Model Law (AML). The OAU has recommended that African countries should include it in their national processes, however, the AML is still being much debated among the various stakeholders and it has not yet been adopted anywhere. In Ethiopia the lack of policy development for pastoral livestock breeding and production is particularly evident. Policy development in the past was characterised by political upheavals, marginalised pastoral populations and unsatisfactory policy interventions (Moris, 1999). Pastoral livestock breeding was interfered with by sedentarisation programmes, establishment of ranches and large scale grazing blocks, subsidised veterinary campaigns, forced user co-operatives and nationalisation of livestock trade. The current government is attempting to decentralise power and implement a federal system. Following the structural adjustment programme the livestock sector was economically liberalised, including privatisation of livestock, marketing, and removal of subsidies on livestock inputs. An encompassing land use policy was not implemented. Infrastructure in pastoral areas is deficient in rural education, pastoral services, marketing and processing facilities. Civic representation is comparatively weak and NGOs are compelled to conform to government priorities. A breeding policy to improve and conserve Animal Genetic Resources is not institutionalised. Government programmes have so far favoured exotic species to increase production in the densely populated highlands, while the lowland regions were neglected. Ethiopia has not yet submitted a National Biodiversity Strategy and Action Plan to the Convention on Biological Diversity (CBD). An Animal Genetic Resource programme is, however, being developed under the Institute of Biodiversity Conservation and Research, which was established in 1998. Aklilu et al. (2002) described the current trade policy development in Ethiopia as reluctant, and the emerging private livestock marketing sector lacks financial capacity. Although livestock movement permits are not required within the country, livestock trade within Ethiopia appears to be subject to heavy taxes. Livestock taxes and transit fees are not standardised, but are collected at locally varying rates. Taxes are not used to improve the efficiency of livestock markets, but contribute to funding regional government offices. The National Bank requires all legal exports to be conducted through a letter of credit. Under this arrangement, exporters of live animals are liable to pay hard currency to the Bank if the importing country should reject the animals for any reason. Under these conditions, the estimated official off-take for cattle is only 8% and falling (FAO, 1999) and is one of the lowest in Africa. Cross border trade, however, has developed remarkably, as pastoralists are attracted by markets particularly in Saudi Arabia, often via Somalia. When privatising the livestock marketing, most of the government owned ranches, abattoirs and tanneries were sold and the government owned Livestock and Meat Development Enterprise (LMDE) was dissolved. In 1998 the Livestock Marketing Authority (LMA) was created under the Ministry of Trade. The LMA aims to promote domestic and export markets. It initiates policies, laws and regulations, issues quality control directives on exportable and importable materials, establishes staging points and quarantine stations for domestic and export trade, promotes organised livestock

424 CURRENT STATUS AND BREEDING ORGANISATION

markets, abattoirs, skins and hides sheds and encourages research on livestock marketing (Negarit Gazeta, Proclamation No 117/1998). However, the LMA’s potential seems to be still constrained as responsibilities have not being clearly assigned between federal and regional administrative structures and between LMA and the veterinary department of the Federal Ministry of Agriculture. For Ethiopia it can be concluded, that the lack of breeding and trade policy favoured uncontrolled crossbreeding, particularly through regional integration programmes and drought rehabilitation measures. Lack of land use policy has indirectly caused a loss of habitat for the indigenous Boran cattle. Inconsistent trade policies hamper making the most of Ethiopia’s livestock production potential, especially live cattle exports. In Kenya the policy situation appears to be more advanced than in Ethiopia, although pastoral development has been neglected in the past and infrastructures are considered insufficient too. Kenyan pastoralists are comparatively better served by rural education and seem to be more aware of commercial livestock production. Regional governments are relatively stable and more services are also provided by NGOs, who are encouraged to facilitate pastoral development. While technical interventions in the rangeland sector have in the past often not been appropriate, more localised and community based initiatives emerge since the 1990s. Kenya also followed the privatisation policy and removal of government subsidies (Moris, 1999). Livestock breeding and trade policies in Kenya are also deficient. Livestock breeding programmes to upgrade local populations with high yielding exotic breeds are said to have had a negative impact on cattle genetic diversity. However, as these crossbreds turned out to be more productive in the highland areas than both purebred exotic and purebred local animals, they were popular among the highland cattle producers, who sought to increase their revenue from cattle keeping. Lowland pastoral and agro- pastoral systems were also targeted by crossbreeding programmes although these mainly focused on higher potential areas (Wanyama, 2003). Kenya’s National Biodiversity Strategy and Action Plan (NBSAP, 1998) recognises the value of AnGR for food security, and the fact that they are being endangered by crossbreeding with exotic germplasm. It states a lack of clear animal breeding policies to promote improvement and conservation of AnGR. Livestock movements are restricted within Kenya. The Kenyan Animal Diseases Act, CAP 364, 1972, which was revised in 1989, determines the declaration of infected areas, the provisions for infected areas (isolation, disinfection and movement of animals) and controlled movement and import of animals (Incorporating Rules L.N.106/1965). The legislation prescribes examining imported animals, issuing health clearance certificates for imported animals, animal health tests and quarantine procedures. The Kenya Stock Traders Licensing Act of 1962, which was revised in 1983, prescribes licenses for trade or barter of livestock, without which livestock are not permitted to travel from high disease risk areas to terminal markets. Provincial and District Commissioners issue stock traders licenses and police, administration police, veterinary officers or inspectors inspect the license on demand.

CURRENT STATUS AND BREEDING ORGANISATION 425

Under Kenya’s Animal Diseases Act and the current livestock disease threats of CBPP6, Rinderpest and FMD7 from the lowlands, access to the terminal livestock market in Nairobi for lowland pastoralists and agro-pastoralists is poor. It is further limited by weak marketing infrastructure such as fragmented stock routes, unusable holding grounds and bad roads. Barrett et al. (2002) criticise the Government’s animal disease quarantines as instruments to protect highlands around Nairobi from competition from pastoral suppliers, thus impeding trade and causing substantial revenue losses for pastoralists. According to Aklilu et al. (2002) the disease free zones from which livestock and livestock products were exported are outdated, quarantine facilities are not fully operational and the rising cross-border livestock imports are not considered. Exports of livestock and livestock products are, however, done through established veterinary procedures. There is a growing interest in promoting livestock marketing both in quality and volume but there is no specific policy to support and facilitate livestock trade. Kenya has imported twice the volume of red meat and offal it has exported between 1996 and 2000. The contributions from cross-border trade with Ethiopia increased, with currently 46% live animal supply from pastoral areas within Kenya, and 26% supply from cross-border trade. The southern African countries of the SADC region have joined in a regional programme to characterise and conserve indigenous breeds, including capacity building and policy formulation (Köhler Rollefson, 2004). Since 1995 all SADC countries have signed the CBD, however, incorporating the international agreement into national policies remains deficient. Moyo (2003) stated that all member countries have zoo-sanitary measures in place (imports, exports, movement), but only few have animal health acts and regulations with specific provisions for AnGR. However, some countries implemented animal improvement acts, and biodiversity strategies are in process. Import and export regulations need to be revised to minimise indiscriminate imports of exotic breeds and regulate the export of indigenous germplasm. There is still a lack of animal identification, recording and performance testing systems. Zimbabwe is in the process of developing a National Livestock Policy. South Africa has strengthened breeding societies as a tool to highlight the importance of breeds and to improve capacity building among breeders. The Livestock Improvement Act (1977) formally recognised breeding societies, with 61 breeding societies in 2001. For 48 further breeds no societies exist, but the breeds are registered in the South African Stud Book. Campher (2003), a businessman from South Africa, analysed the constraints in international trade in animal genetic material. Accordingly, the trade of commercial animals from other countries into the SADC region as well as within SADC-countries is common. After lifting the sanctions to South Africa in the early 1990s veterinary protocols were established for exporting semen, embryos and live animals. In the absence of further regulations, breeders negotiated on their own with foreign buyers. The demand for genetic material from South Africa was initially high, but decreased

6 CBPP: Contagious Bovine Pleuro-Pneumonia. 7 FMD: Foot and Mouth Disease.

426 CURRENT STATUS AND BREEDING ORGANISATION due to lack of performance control of the exported animals and lack of standardised protocols in recipient countries. Efforts to form trade organisations, such as the South Africa Australia Breeders Company (SAABCO) failed due to lack of experience.

WORLDWIDE TRANSFERS OF BORAN AND TULI LIVE CATTLE, 427 SEMEN AND EMBRYOS

4 WORLDWIDE TRANSFERS OF BORAN AND TULI LIVE CATTLE, SEMEN AND EMBRYOS The worldwide transfer of Boran and Tuli live cattle, semen and embryos (cf. Table 3) started in the late 1940s. It was prompted by the research and commercial interests of European settlers who traded live cattle in different African countries. After recognising the potential of both breeds for beef production under environmental constraints in the late 1980s, Boran and Tuli germplasm was exported from Africa, mainly through the Australian partnership between research and the beef industry. The transfer was then extended to North and South America. Since the 1990s the two breeds have been used in research and crossbreeding programmes run by the world’s major beef producing nations. Tuli cattle have gained appreciation as a promising breed for quality beef production, and the Boran mainly for improving fertility and adaptation traits in international beef breeds. Nevertheless their proliferation in the beef industry seems to be overestimated. In order to describe the transfers of Boran and Tuli cattle breeding material from their countries of origin, data were collected from the literature, the internet, and by contacting research, development and business experts with a stake in cattle breeding worldwide. Key persons involved in the transfer of Boran and Tuli breeding material were identified and contacted. The lack of documented data on cattle numbers, prices and agreements limited the supply of concrete data. The objective was therefore reformed to appraise rapidly the origin, major transfers and key investors in the Boran and Tuli cattle genetic material where data were available.

4.1 Transfers within Africa Kenya has exported Improved Boran live animals to several African countries, and embryos to South Africa. Data on the transfers could neither be obtained from ILRI, nor from the Boran Cattle Breeders Society. Further correspondence was with most of the Boran breed importing countries, but very little data was returned. Only Embryo Plus, a private company in South Africa, shared their information on embryo transfer. The sources were therefore mostly scientific literature. No information was obtained on Boran cattle in Congo. In 1947 Kenyan ranchers transported live cattle by ship, rail and on the hoof to northern Zambia. In the following 30 years more best quality Improved Boran were exported from Kenya to the commercial ranch sector in Zambia, which provided good management conditions (Frisch, 1989). In 1965 the Zambian Department of Agriculture started beef cattle research at the Central Research Station, Mazabuka, testing the Boran as sire and dam breed (Thorpe and Cruickshank, 1980). The aim was to evaluate the beef potential of major cattle breeds under ranching conditions. Anticipating a continued development of the beef industry, the research programme established foundation breeding herds of purebred dams. 13 Boran bulls sired progeny during the programme. According to Frisch (1989) the breed prospered in Zambia and the Zambian Boran Breeder Society had the largest number of registered cattle of any breed in Zambia.

428 WORLDWIDE TRANSFERS OF BORAN AND TULI LIVE CATTLE, SEMEN AND EMBRYOS

First imports of Boran cattle into Tanzania may have occurred in the late 1940s when a Boran dairy herd was established by the Government in Tanga and when some white settlers in the Arusha/Moshi region bought Boran cattle for beef breeding. During the 1950s and early 1960s parastatal and company ranches were the main importers of Improved Boran, especially bulls to upgrade small local zebu cattle (Meyn, 1970). In Uganda, research on hybrid vigour of Boran cattle was started at Ruhengere Field Station in the late 1960s, using four Boran bulls for crossbreeding beef cattle. First results revealed the importance of hybrid vigour for beef cattle, with Boran cows weaning heavier calves and showing good mothering ability. Larger numbers of Boran purebreds were produced in 1969 to be used as sire breeds for crisscrossing on west Uganda ranches (Sacker et al., 1971). Recently, the privately owned Zziwa ranch - formerly property of the Uganda government ranching scheme - claims to possess 3,000 best quality Boran cattle, and considers these as the best of all local breeds. The ranch is the leading provider of semen in Uganda and supplies large scale beef producers. It purchases semen from South Africa and does artificial insemination. Some of the animals are also bred for beef production, with leading beef dealers as main customers. It also exports live bulls and beef to Kenya and European countries (All.Africa.com, 2005). De la Ray (2005, pers. comm.) reported a substantial movement of about 200 Boran bulls per year from South Africa to Uganda over the last five years. The imports are driven by the Ugandan government with international funding. Improved Boran cattle are reported to be widely distributed in southern Nigeria (Nwosu et al., 1985). They were imported as heifers from Kenya in 1975. Since then they were kept as a closed herd at the University of Nigeria Teaching and Research Farm Nsukka, southern Nigeria, to improve beef production. They proved susceptible to trypanosomes, but tolerated the disease under regular treatment. The cows had a high milk output and raised heavy calves under the Nsukka farm conditions. The average body weights were comparatively heavy for Nigerian cattle (345 kg + 10.41), but were below the Kenyan Improved Boran weights. In South Africa, Embryo Plus, a private company, was the main importer of Boran germplasm. They first transferred about 600 Boran embryos from Kenya to Zimbabwe in 1992. They imported 481 (1994), 365 (2000), 981 (2001), 603 (2002) and 331 (2004) Boran embryos from Kenya to South Africa. Further, they imported 6 cattle from Zimbabwe in 2003 and 10 cattle from Swaziland in 2004; and they transferred 120 cattle from Zambia to Swaziland in 2003 (de la Ray, 2005, pers. comm.). Pretorius (2004) adds that 60 cows, 30 calves and seven bulls were imported from Zambia in 2004. After their potential for commercial beef production was recognised in Zimbabwe, Tuli cattle were transferred to the neighbouring African countries. However, access to information on Tuli transfers seemed even more difficult than for the Boran, probably as the Tuli population is comparatively small. No response was obtained from any of the Tuli Breeders Societies. Contacts to Zimbabwe confirmed the available information but did not specify the size and payments for Tuli exports. Information

WORLDWIDE TRANSFERS OF BORAN AND TULI LIVE CATTLE, 429 SEMEN AND EMBRYOS

from Botswana was limited to scientific literature. In South Africa Embryo Plus provided the only information on Tuli transfers. No information was obtained for Namibia and Gabon. The first Tuli cattle were imported to South Africa from Zimbabwe in 1977. Regular imports of superior cattle from Zimbabwe are reported to have broadened the genetic base, but numbers were not available. South Africa claims now to have a sufficient gene pool to meet future breeding requirements (TCBS, 2004). During the 1990s several auctions were held by the Zimbabwean and South African Breeder Societies and approximately 200 of the best Zimbabwean Tuli cattle were transferred to South Africa. 13 of Zimbabwe’s best Tuli bloodlines are now available in South Africa (Pretorius, 2004). Live Tuli cattle of the improved type were (re-)exported to Botswana in the past, but the date and investigators could not be specified. According to Mpofu (2002), the Botswana Government owned a Tuli herd of 57 cows in 2000 and used them for breed evaluation studies.

4.2 Import into Australia and America In the 1980s and early 1990s, both Australian and USA beef industries developed rapidly, receiving a particular push from research and technology development in breeding, feeding and marketing. Their beef sectors opened to the world market and hence underwent increasing pressure to import and exploit new breeds, in order to match cattle traits with market and consumer requirements, and to improve beef quality. As there were not yet any local breeds other than the Brahman8 that were adapted to environmental stress with good productive traits, the Boran and Tuli were hailed by some as new options to increase the productivity of composite crossbred populations due to their additive contributions and heterosis effects (Vercoe, 1989; Burrow et al., 2001; Lunstra and Cundiff, 2003).

4.2.1 Australia Documentation of the transfer of Boran and Tuli genetic material to Australia and of their dispersal in the Australian beef industry is scarce. Vercoe (2004, pers. comm.), who was involved in the initial transfer of Boran and Tuli embryos, kindly provided information on the volume and the mode of transfer. The Commonwealth Scientific and Industrial Research Organisation (CSIRO) archive statistics could not be accessed. Individual beef producers kindly provided case-to-case information, but could not provide aggregated data on the volume of transfers. No reply was received from the Australian Boran and Tuli cattle breeders associations. A descriptive account of the Boran and Tuli transfers from Africa to Australia was obtained only from a CSIRO perspective, but not from African counterparts. The first scientific interest in Australia in crossbreeding with zebu cattle at the end of the 19th century was opposed by the Industrial United Graziers Association, due to

8 The Brahman is a Bos indicus breed developed in the USA from Indian breeds. It is popular for its good adaptation to tropical and subtropical conditions, desirable maternal traits, calving ease, and high levels of heterosis in crosses with Bos taurus breeds (Ralph, 1992).

430 WORLDWIDE TRANSFERS OF BORAN AND TULI LIVE CATTLE, SEMEN AND EMBRYOS

their prejudice against non-British cattle. However, in 1933 the CSIRO forerunner, the Council for Scientific and Industrial Research, co-operated with a group of Queensland graziers to import the first Brahman cattle from the USA. These were, however, not sold or distributed for breeding until 1942. In 1953, initiated by the United Graziers Association, the Australian Meat Company Board purchased two properties for research in Queensland. CSIRO operated research at the National Cattle Breeding Station, Belmont, and their results supported the observation of the beef industry, that Brahman bulls used on British cows resulted in superior crossbreds. However, after an additional import of Brahman in 1958 a complete ban was issued on live cattle imports to protect the cattle industry from accidental diseases. In 1986 CSIRO, with support from the Australian Centre for International Agricultural Research (ACIAR), had reviewed African cattle breeds, and the Tuli and Boran breeds appeared most promising. In consequence, Boran and Tuli embryos were imported into Australia, initially for research only (Ralph, 1992). The arrangements for the import of deep frozen Boran and Tuli cattle embryos from Africa into Australia were made in 1987. According to Vercoe (1992) the Australian Quarantine and Inspection Service developed protocols and the Australian policy bodies provided endorsement. Frisch (1989) stated further, that the initial negotiations were made by CSIRO representatives with the Boran Cattle Breeders Society in Kenya, the Tuli Cattle Breeders Society in Zimbabwe, and later also the Boran Cattle Breeders Society Zambia. While the breed societies were very enthusiastic to enter into a co-operation, the Kenyan and Zambian authorities were hesitant. After 12 months of negotiation the Kenyan and Zambian authorities agreed to the transfer. Thereafter, CSIRO’s approach to the Tuli Cattle Breeder Society in Zimbabwe was also successful. However, the contents of the negotiations between CSIRO and the African governments, their scepticism and motivation to agree, were not specified. Finally, CSIRO decided to collect Boran embryos from Zambia, and not from Kenya due to political reasons. Tuli embryos were collected from Zimbabwe. The implementation of the embryo transfer programme started in 1988. Frisch (1989) reported that the potential Boran and Tuli donor cows were selected by the Boran and Tuli Cattle Breeding Societies in Zambia and Zimbabwe. After passing foot and mouth disease tests, about 20 bulls and 80 cows of each breed entered the final quarantine phase after three months at stations in both countries. The Zambian Department of Veterinary Services and Tse-tse Fly Control made their quarantine station available, which was close to the capital Lusaka. The station required considerable upgrading to meet Australian quarantine requirements. After successful embryo collection in Zambia, the CSIRO team departed for Zimbabwe to collect Tuli embryos. In Zimbabwe a private company had constructed quarantine facilities, and had already collected sufficient Tuli semen and placed the Tuli cows in quarantine before the CSIRO team arrived. According to Vercoe (2005, pers. comm.) the Australians were better served in Zimbabwe due to the presence of an independent veterinarian who collaborated with the Zimbabwean government. In 1988, CSIRO in collaboration with the Boran and Tuli Producer Consortium imported the first deep frozen Boran (264) and Tuli (269) embryos to Australia. This

WORLDWIDE TRANSFERS OF BORAN AND TULI LIVE CATTLE, 431 SEMEN AND EMBRYOS was the start of a large scale practical test of embryo transfer for cattle breeding. The embryos were transported via air to the Australian off-shore Quarantine Station on Cocos Island, where they were transferred into Australian recipient cows. Several hundred Friesian heifers had been tested on-farm for diseases and 112 were considered suitable to act as surrogate mothers. They were flown there specifically for this purpose and 93 cows were registered pregnant from 12 sire lines (Frisch, 1989). In 1990, 73 Boran and Tuli calves of similar size were transferred by air from Cocos Island and 71 arrived safely in Australia after six months of extensive tests for diseases. Some were kept as bulls in a tick and blue-tongue free quarantine in southern Australia for future artificial insemination. The others were shipped to Belmont in Northern Australia, the CSIRO National Cattle Breeding Station near Rockhampton, Queensland (Vercoe, 2004, pers. comm.). Taylor (1994) reported that in Belmont extensive crossbreeding programmes were launched with about 500 cows. Several thousand cows were involved in crossbreeding programmes of the Boran and Tuli Producer Consortium members. Young cattle were grown out on Belmont or nearby properties and the heifers were subjected to embryo transfer procedures in an effort to multiply the number quickly and to get them out to the industry via the consortium. VAB Genetics designed artificial insemination techniques to assist the rapid build-up of pure and crossbred herds with Boran and Tuli semen. The crossbreeding programmes served to evaluate the Boran and Tuli breeds for various production, reproduction and health traits on pasture. Results of crossing Adaptaur and Belmont Red cows with Boran, Tuli and Brahman sires showed that crossbred cows were more productive than purebreds. The F1 crosses with Boran and Tuli had higher calving rates than the purebreds and F1 crosses with Brahman. Survival of Boran and Tuli crossbreds to the age of 18 months was higher than for the purebreds. Boran and Tuli sired cows were constantly more productive than the larger Brahman crosses, although they had progeny with lower mature live weights than Brahman sired cows (Gazzola et al., 1998). Frisch and O’Neill (1998) found no difference in tick resistance between different breeds and crosses. Boran cattle were slightly more resistant to ticks than Tuli, but could not equal the Brahman for resistance to both ticks and worms. Boran sired progeny as well as Brahman sired progeny had higher live-weight gains than purebreds irrespective of parasite challenge. Tuli sired progeny exceeded the purebreds only at lower parasite level. Burrow et al. (2001) concluded that as the environmental conditions became harsher, better adapted breeds produce relatively better quality meat than less well adapted breeds. In 1993, CSIRO and the consortium held the first auction for 40 purebred Boran and Tuli embryos, the progeny of the original import in 1990. The auction attracted interest from areas as diverse as Western Australia, Victoria and Tasmania (Taylor, 1994). The second Boran and Tuli purebred embryo sale was held in 1994. The Tuli embryos were in demand and fetched a new world record price of US$ 9,500. The top price paid for the Boran embryos was US$ 5,000. In 1995 the original purebred herd of imported Boran and Tuli cattle and their progeny were further dispersed to the industry, and all progeny of the original 1990 CSIRO import were put on the market

432 WORLDWIDE TRANSFERS OF BORAN AND TULI LIVE CATTLE, SEMEN AND EMBRYOS

(Temere Pastoral, 2005). The CSIRO corporate press release 95/32 quotes Vercoe saying that marketing embryos through auctions speeds up the development of the breeds considerably, as the producers are more competitive and effective in tuning the breeds to suit industry needs. The whole import programme of Boran and Tuli embryos was described as a very successful venture between scientists and the Australian beef industry. However, despite the promising research results and the enthusiasm of the consortium, both breeds have not been largely accepted by the Australian beef industry. According to a later statement by Vercoe (2004, pers. comm.), Tuli were incorporated mainly into a few composite herds and were blended with the Belmont Red, Bonsmara and Senepol. Several purebred breeders provided the gene pool and only one of the large cattle companies made extensive use of the Tuli after the results of privately imported Tuli cattle became visible. The Boran in the end did not compete favourably with the Brahman and only a few breeders incorporated it into composites. It seemed that the comparatively lower body weight was the most striking argument against preferring Boran and Tuli over Brahman. The Australian Tuli Association was founded in 1990 and has currently 28 members (Agfact, 2004). The Association has set up a Tuli Group BREEDPLAN marketed by the Australian Business Research Institute (ABRI) with an integrated performance and pedigree recording (Temere Pastoral, 2005). About twenty breed societies have genetic evaluations on their databases. The Boran Association was founded in 1992, but information about their current operation could not be found. Recently the Woorabinda Pastoral Company was founded as a pilot project under a draft strategy to evaluate the Boran and Tuli breeds for fast growing grass-fed cattle suitable to meet the demands of the Asian market. This project was managed with CSIRO and is the first independent large scale commercial evaluation of Boran and Tuli cattle in Australia. It is designed to provide breeding information to other pastoralists, particularly Aboriginal enterprises, and to supply bulls commercially (ATSIC, 2004). However, according to Allen (2001, compiled by the Australian Registered Cattle Breeders’ Association Inc.), Australia had 60 registered Tuli cattle in 1995 and 161 in 2000. The registered Boran cattle were 41 in 1995 but only 26 in 2000. Given the size of the overall beef industry with 26.4 million head of beef cattle in Australia in 2004 (FAO STAT, 2005), the population of both breeds appears comparatively insignificant.

4.2.2 America Boran and Tuli breeding material has entered North, Central and South America since 1991. Sources of information in America were difficult to access. Therefore key persons in Australia and South Africa, as well as the available scientific literature and media releases were drawn upon. The USA launched the largest body of research on the Boran and Tuli breeds so far in 1991. Due to health restrictions, they did not import genetic material directly from Africa, but started by importing semen from Australia. The joint venture of CSIRO and the Boran and Tuli producer consortium sold approximately 250 doses of semen from eight Boran and nine Tuli bulls to the US Meat Animal Research Centre (MARC) in Nebraska, for testing in the germplasm evaluation programme (Cundiff et

WORLDWIDE TRANSFERS OF BORAN AND TULI LIVE CATTLE, 433 SEMEN AND EMBRYOS al., 2000). MARC cooperated with the Texas A&M University Research Centre at their subtropical research stations in Texas, Florida, Georgia, Louisiana, and Oklahoma (Hill, 1993; Herring et al., 1996; Cundiff et al., 2000). At the Nebraska site, the germplasm evaluation programme primarily studied carcass characteristics and growth of Boran and Tuli compared to other breeds. Early results showed that Tuli produced crossbred progeny with both carcass and meat quality similar to Charolais, Hereford and Angus and better than Bos indicus breeds. Boran beef showed more variation but was more tender than beef from Brahman. Herring et al. (1996) showed that Tuli sired carcasses had advantages in value based marketing systems, and that the smaller body size of Tuli and Boran sired females bore lower maintenance costs. Boran and Tuli crosses had relatively low average daily gains but were younger at puberty than Brahman sired crosses. Brahman sired crosses had higher maternal weights than Boran, and Boran had higher weights than Tuli (Cundiff et al., 1996). Lunstra and Cundiff (2003) found that bulls sired by European breeds grew more rapidly than bulls sired by Boran and Tuli breeds. In the light of the high conception rates and the high birth and weaning percentages of the Tuli crosses, Warrington et al. (2004) concluded that Tuli crosses would have the greatest overall reproductive efficiency under the semi-arid conditions of south Texas. Cundiff et al. (2000) suggested that Tuli might replace Bos indicus breeding and maintain adaptation traits without reducing meat tenderness. The Tuli Association captured these results to praise the tenderness of Tuli beef as its highest inheritable trait, arguing that the industry needs to move towards a moderately sized animal to raise beef quality (Briggs, 2002). Since then various industrialists have imported Tuli cattle into the USA from Australia and Zimbabwe via Canada. Maynard (2005, pers. comm.) mentioned several players in Australia who exported Boran and Tuli genetic material to America. Leachman Cattle Co., USA, bought Tuli semen from Australia and sent some further on to South America. RAB-Australia (a privately owned artificial breeding organisation) used to send Tuli semen to their counterpart RAB-Mexico. Teeraweena Boran and Tuli was very active in exporting genetic material to south America. Unfortunately none of the companies provided further information. According to a Dallas Business Journal media release in 2004, the industrial germplasm producer Carrol Shelby formed a syndicate to transfer Tuli embryos. He purchased the first Tuli from his partner Scott McKay, Calberta Farms and Northern Vision Co., Canada. McKay had travelled to Zimbabwe to select high quality Tuli cattle and shipped embryos to Canada, where they were placed in recipient cows. Out of these, Shelby purchased the top five purebred Tuli yearling heifers and one purebred Tuli bull and took them to Texas. Today, Shelby keeps about 700 Tuli cattle and sells Tuli semen to several cattle raisers in Mexico and South America. Shelby’s partner Ken Briggs had started the first Tuli herd in the USA in 1996. Briggs specialises in top grade beef. He has about 275 head of Tuli cattle, which he crossbreeds with Japanese Wagyu cattle to develop a composite breed under the trademark ‘Waguli’. A Country World (2005) media release features Bill Bucek, another partner of Shelby, who introduced the first live Tuli bull from Australia into the United States in 1995. Convinced by a business associate in Australia, he brought

434 WORLDWIDE TRANSFERS OF BORAN AND TULI LIVE CATTLE, SEMEN AND EMBRYOS the bull in after a two-year quarantine period. Bucek breeds for meat quality. He also developed a composite breed named ‘Tulangus’ by crossing 40 registered Red Angus cows with Tuli bulls. The Texas A&M research stations in Texas also collaborated with the Mexican National Institute for Agricultural Research (INIFAP) and the Foundation for Livestock Research in the State of Sonora (PATROCIPES). They pursued research in four Mexican states since the late 1990s, focussing on crossbreeding schemes with artificial insemination using Tuli semen on local cows. Reproductive performance, adaptability, and carcass characteristics were evaluated with a view to producing desirable characteristics for re-export to the USA (Fajerson, et al., 1996). The North American Tuli Association (NATA) was formed in 1998 with a board of directors from the USA, Canada and Mexico. Currently there are about 26 breeders associated with NATA. In 2002 NATA sponsored a Tuli field day at a private ranch in Texas (Briskin, 2002). The Texas Department of Agriculture granted the North American Tuli Association US$ 20,000 for promoting Tuli cattle among breeders throughout the USA and Mexico (TDA, 2004). According to TDA (2004), 90% of the purebred Tuli cattle in the USA were reported in Texas, with about 500 registered cows (Briskin, 2002), and about half of these kept by Shelby (cf. above). For the Boran population no figures were found. Hammack (2003) classified the Tuli population as less significant in the Texas region and did not mention the Boran at all. Again, given the size of the overall beef industry with 94.9 million head of beef cattle in the USA (FAO STAT, 2005), the population of both breeds appears insignificant. Apart from the USA, Tuli embryos were imported from South Africa to Canada and Latin America. Embryo Plus exported 450 embryos to Canada in 1995 and 189 embryos to Argentina in 1996, but no legal exports of live cattle or semen were recorded (de la Ray, 2005, pers. comm.).

WORLDWIDE TRANSFERS OF BORAN AND TULI LIVE CATTLE, 435 SEMEN AND EMBRYOS

4.3 Diagrammes of Boran and Tuli transfers

USA

Nigeria Ethiopia Mexico Somalia

Brazil Uganda Kenya Eastern Africa North- and South America Congo Tanzania

Zambia Zimbabwe

South Africa Swaziland

Unimproved Boran Australia

Improved Boran Australia Southern Africa

Figure 2: Worldwide transfers of Unimproved and Improved Boran cattle breeding material from Eastern and Southern Africa

North- and South America Canada

USA

Mexico Gabon Australia Argentina Australia

Namibia Botswana Zimbabwe

Southern Africa South Africa Tuli

Figure 3: Worldwide transfers of Tuli cattle breeding material from Southern Africa

Table 3: Transfers of Improved Boran and Tuli cattle breeding material from eastern and southern Africa

Country Date of Breeding material Purpose Type of Way of Importer Exporter Cryo- Actual Further import transfer transfer conser- stock distribution vation numbers Improved Boran in the African region Zambia 1947 Live animals, Improved Commercial Ship, rail, on Commercial Commercial Yes < 10,000 30 years number unknown beef the hoof ranch ranch in Kenya Î ongoing production Tanzania Late Live animals, Dairy Not known Not known Government Not known < 5,000 Ð 1940s number unknown production

Improved Commercial Not known beef ranches

production

1950/ Live animals, Upgrading Parastatal and Not known early mainly bulls of small commercial 1960s zebu cattle ranches

Uganda 1960s Live animals, Improved Research and Not known State research Not known Not known Beef and live number unknown beef and commercial station bull exports to germplasm Kenya and 1969 Semen, production Commercial South Africa Europe amount unknown ranch Since + 200 bulls year-1 Government South Africa 1999 with int. funding Congo Not Not known known Nigeria 1975 Live animals, Improved Research Not known State research Not known Not known number unknown beef station production

Table 3 (continued) Country Date of Breeding material Purpose Type of Way of Importer Exporter Cryo- Actual Further import transfer transfer conser- stock distribution vation numbers South 1994 481 embryos Improved Commercial Not known Embryo Plus, Kenya 61 USA, Canada, Africa (1995) not known beef and South Africa Zambia registered Argentinia 2000 365 embryos germplasm Kenya (1998) Ï 2002 981 embryos production Kenya 2003 603 embryos Kenya 6 cattle Zimbabwe 2004 331 embryos Kenya 10 cattle Swasiland 60 cows Not known Zambia 30 calves 7 bulls Zimbabwe 1992 +600 embryos Improved Not known Not known Embryo Plus, Kenya Not known beef South Africa production Swasiland 2003 120 cattle Improved Not known Not known Embryo Plus, Zambia Not known beef South Africa production Tuli in the African region Botswana Not Living animals, Improved Research and Land State research Yes 57cows at known number unknown beef commercial station (1990) study production (2000), population data not available Ð Gabon Not Not known known Namibia Not Not known known

Table 3 (continued) Country Date of Breeding material Purpose Type of Way of Importer Exporter Cryo- Actual Further import transfer transfer conser- stock distribution vation numbers South 1977 Many cattle Improved Research and Auctions, SA Breeder Zimbabwe Yes 4,000 ???? Africa ongoing beef commercial land Societies Breeder Societies registered 1999 200 living animals production (2004) Boran and Tuli in Australia and Northern and Southern America Australia 1988 20 bulls and 80 Improved Research and Plane CSIRO and a Zambia and 26 Canada, USA, cows of Boran and beef and commercial producer Zimbabwe registered South Africa Tuli each germplasm consortium Boran 1991 264 Boran embryos production (2000), 269 Tuli embryos 161 registered Tuli (2000) USA 1991 250 doses semen Improved Research and Plane ÚS MARC CSIRO and a 1,500 Mexico and from 8 Boran and 9 beef and commercial producer registered South America Tuli bulls germplasm consortium, Tuli (2004) production Australia Not 5 Tuli heiffers, Germplasm Not known Carrol Shelby Scott McKay, known 1 Tuli bull and Calberta Farms improved and Northern beef Vision Co., production Canada 1995 1 Tuli bull Improved Not known Bill Buceck Australian partner beef production 1996 Not known Improved Not known Ken Briggs Not known beef production Not Tuli semen Not known Not known Leachmann Australia known Cattle Co.

Table 3 (continued) Country Date of Breeding material Purpose Type of Way of Importer Exporter Cryo- Actual Further import transfer transfer conser- stock distribution vation numbers Canada Not Not known Improved Commercial Ship Scott McKay, Zimbabwe Not known USA known beef Calberta production Farms and Northern Vision Co. 1995 450 Tuli embryos Not known Embryo Plus, South Africa Argentina 1996 189 embryos Not known Commercial Not known Not known Embryo Plus, Not known South Africa Mexico Not Tuli semen Not known Commercial Not known RAB Mexico RAB Australia Not known known

440 MODE OF TRANSFER

5 MODE OF TRANSFER Mode of transferring genetic material in subsistence systems traditionally refers to herd migrations, exchange mechanisms within and between livestock keeper communities, and local trade with live animals. In commercial systems it traditionally refers to transporting live animals. New transfer mechanisms include advanced breeding organisations and breeding strategies, embryo transfer and artificial insemination, vertical integration of cattle breeders with livestock research and the beef industry, and horizontal integration of peer groups. Progress in breeding and transfer technology has widened possible genetic transfers and sped dispersion. With regard to Boran and Tuli genetic transfers, NGOs have staged a debate on access and benefit sharing and advocate agreements to be implemented. The following section attempts to review the different mechanisms behind the worldwide transfer of Boran and Tuli cattle genetic material. The initial joint venture of CSIRO and the Australian beef producers to import Boran and Tuli embryos from Zimbabwe and Zambia became a prominent issue in the public debate and information was available from several sources. However, detailed information is scarce on other transfers to industrial beef producers and possibly other research institutions in Australia, North and Latin America, and on the later expansion of trade within Africa and worldwide. Data on the volume of transfers were usually unavailable and no references were found to payments, be it for lack of access to the appropriate sources, or for lack of documentation. The section is therefore limited to deriving preliminary assertions.

5.1 Traditional and modern transfer mechanisms Since historical times the distribution of cattle genetic diversity in sub-Saharan Africa has been largely through human migration and was influenced by adaptation to specific habitats and diseases (Hanotte et al., 2000). Breeders’ selection among the small shorthorned zebu cattle introduced via Saudi Arabia was predominantly for adaptive traits and this permitted their long migrations from East Africa to southern Africa and their fast adaptation to new environments. Traditional exchange mechanisms, local trade among and between different ethnic groups, raiding and warfare further dispersed the genetic material at a relatively low speed. The systematic improvement of the Boran and Tuli genetic material - including its transfer - and the formation of breed organisations was stimulated differently for the two breeds. For the Boran, it was triggered by commercial interests of European settlers in Kenya and by an increasing demand for beef and dairy cattle. Since the 1920s, settlers in Kenya purchased Boran cattle originating in southern Ethiopia / northern Kenya mainly from Somali traders but probably also from local pastoralists. Within breed selection under improved management and in prime grazing areas in Kenya rapidly developed the Improved Boran as an excellent beef breed. The foundation of the BCBS in 1951 has advanced professional testing and evaluation of the breed and was the starting point for establishing a national beef cattle breeding scheme. Although accompanied by extensive research programmes launched by KARI, this scheme was mainly borne by the commercial interest of private ranches and led to the Improved Boran. Buyers of the animals were Kenyan government agencies co- operating with research and breeding institutions. Sellers were the commercial ranchers and/or pastoralists. The commercial breeders’ investment has contributed to maintain and to propagate the Improved Boran (Mason and Maule, 1960; Hetzel, 1988; Felius, 1995).

MODE OF TRANSFER 441

The development of the Tuli breed in Zimbabwe is founded on a government intervention in the early 1940s, when it established the Tuli Breeding Station. The Tuli Cattle Breeder Society was founded in 1961 and the use of Tuli cattle increased over years both in the commercial and the smallholder sector. According to Moyo (2005, pers. comm.), the continued demand for Tuli genetic material from inside and outside Zimbabwe kept the breeders in Zimbabwe busy and keen to expand their herds, knowing that there is a ready domestic and international market. With the Boran and Tuli embryo exports to Australia, the dispersal of the breeds gained a new momentum. Australia had the necessary industry, infrastructure and scientific resources to undertake this endeavour, thus creating an efficient and relatively low cost transfer mode, replacing live animal transfers and interrupting disease-transmission cycles. This was the beginning of the breeds’ dispersal to Australia and the Americas. According to Vercoe (1992) the undertaking was a practical approach to genetic improvement in harsh environments, and an opener for a promising new area of molecular genetics, identifying useful genes and combinations to improve production potential and resistance to stress. Embryo transfer and artificial insemination technologies require sophisticated husbandry management, high nutritional standards, disease control and infrastructure, and are thus limited to high input cattle systems. The technology rapidly became a tool for business transactions of Boran and Tuli genetic material between individuals, institutions and companies in Australia, the Americas and South Africa. It thus permitted the beef industry to supply the world market with new cattle genetic material. In Africa, countries like Zambia, Botswana and South Africa made use of cryo-conservation of Boran and Tuli semen. On one hand, these technologies can overcome health constraints and sanitary barriers and reduce the costs otherwise associated with live animal exports. On the other hand, they may also lead to erosion of genetic diversity because genetically similar material from locations where the technology is available is propagated. These technologies may thus contribute to exclude poor livestock keepers from the market, because they do not usually have access to the technologies and are unable to compete (Köhler-Rollefson (2000). The collaborative project between CSIRO, the Boran and Tuli Producer Consortium and cattle breeders is an example for the vertical integration of livestock research, beef industry and breeders in a joint venture. The CSIRO National Cattle Breeding station in Belmont was owned and funded by the Meat Research Corporation. Similar cooperations were found between the Texas University and individual cattle businessmen in the USA. This vertical integration was part of the initial success of the venture. Horizontal integration between breeders’ societies and beef producers across countries is another element to foster joint breed improvement, genetic progress evaluation, and marketing and promotion programmes. In Australia, the Boran and Tuli breeder societies merged their genetic evaluation database into BREEDPLAN, which is marketed by the Agricultural Business Research Institute. The North America Tuli Breed Society has amalgamated the USA, Canada and Mexico for joint breed improvement and promotion programmes. The South African Tuli Breeders Association, including Namibia, has a joint breed evaluation programme with the Tuli Breeders Association in Zimbabwe (BLUP). In eastern Africa similar joint programmes do not seem to be established yet. A further example of joint interest groups are the USA Shelby producer syndicate.

442 MODE OF TRANSFER

5.2 Issue of access and benefit sharing agreements Access and benefit sharing agreements arose as a critical issue to conserve farm animal diversity. Industrial companies along with major governmental and academic research institutions are accused of exploiting the Boran and Tuli cattle breeds - sovereign assets of African countries - and thereby commercially exploiting local breeders’ indigenous knowledge and the genetic resources they developed. It is further argued that the two breeds have added to the world’s beef market and to the creation of valuable new cattle breeds without any benefit being returned to the countries of origin. The issue that the Boran and Tuli genetic material was accessed without an appropriate framework for fair and equitable access and benefit sharing has been raised by NGOs, namely the Canadian ETC group (formerly Rafi) (RAFI, 2000), the Germany based League for Pastoral Peoples (LPP, 2002; Gura, 2003), the Scandinavian based AfriDAD (Zulu, 2003) and New African (Commey, 2003). The discussion staged on access and benefit sharing to date exclusively concentrates on the exports of Boran and Tuli genetic material to countries outside Africa, and still avoids a more differentiated view on the “inner-african” transfers - form the area of origin to other African regions. Boran cattle were transferred from Ethiopia via Somalia and Kenya to Zambia, Tanzania, Uganda, Congo, Nigeria, Zimbabwe, Swaziland and South Africa (cf. Figure 2), where they were used by different stakeholders for commercial purposes. Here again, the benefits accrued from such transfers can hardly be – even roughly – estimated. Nevertheless, if a balanced discussion on access and benefit sharing is to be achieved, transfers of this nature must as well be taken into account. The CBD (1992), which was concluded four years after the start of trading in Boran and Tuli embryos to industrialised countries, is the major legally binding instrument for the trade in genetic resources. It sets out the fair and equitable sharing of the benefits arising from using genetic resources as prime objective (CBD Article 1). With regard to access to genetic resources, Article 15 recognises the sovereign rights of every state and subject of national legislation. Access is to be based on Prior Informed Consent (PIC) and mutually agreed terms (MAT) about the objectives and about economic and environmental implications of such access, granted through bilateral agreements. The FAO Global Strategy for the Management of Farm Animal Resources (1999) further acknowledges that the most rational and sustainable way to conserve animal genetic resources is to ensure that locally adapted breeds remain a functional part of their production system. Therefore the full value of local breeds must be taken into account and the rights of livestock breeders to multiply, exchange and develop breeds must be protected. The apparent lack of prior informed consent is one of the major points of discussion, and the sovereignty and fair trade in Boran and Tuli genetic resources is often questioned by those criticising the venture. Some sources maintain that the initial embryo transfer from Zimbabwe and Zambia to Australia was agreed for scientific purpose in crossbreeding at CSIRO only. However, a second transfer one year after the arrival of the first progeny in Australia was sold by the CSIRO joint venture to the US MARC, and the dispersal of Boran and Tuli cattle to the Australian beef industry started immediately after their arrival in Australia. Individual and company business relations attempted to push the commercial use within Australia and towards Northern and Southern America.

MODE OF TRANSFER 443

RAFI (2000) commented that those involved in the transfer of Tuli cattle concede that the removal was undertaken with as low a profile as possible, and that Zimbabwe officials would have opposed the transfer, had they been aware of the full endeavour. Other sources go as far as accusing the Australians to have stolen the embryos from Zambia and Zimbabwe (Zulu, 2003) and argue that the benefit to the African countries is nil (Commey, 2003). According to Mushita (2005, pers. comm.) there was insufficient transparency about the full commercial use of the genetic material. However, although no formal contract between CSIRO and the Zimbabwe government was made, a certificate was granted to export the embryos from Iridor Farm in Zimbabwe. According to Vercoe (2004, pers. comm.), agreements with the authorities in Zambia and Zimbabwe were in place. The Australia Quarantine Inspection Service was involved in the negotiations and in the development of protocols. Part of the agreements was to upgrade the local quarantine stations and to provide infrastructure, equipment and training, which would enable the local authorities and institutions to pursue their own programme to export cattle embryos. CSIRO established quarantine stations and veterinary facilities in both countries and two qualified veterinarians from each country received a 2 months training programme in embryo collection, evaluation and transfer at Rockhampton, Australia. The facilities were intended to be used after the departure of the Australians. While embryo transfer has since continued at the Zimbabwe station, the infrastructure built in Zambia had gone a year later. The payment of appropriate benefits from the commercial exploitation of Boran and Tuli cattle to their original breeders has become a prominent issue in the debate. However, as the Boran and Tuli were intended and have been used by the consortium mainly for crossbreeding and to form composite breeds, it is difficult to estimate their current value. RAFI (2000), one of the main advocates of such payments to be implemented, have based their calculations on early Australian media releases and suggest a minimum of 5% of the value added by Boran and Tuli genetics to the Australian beef sector to be paid as remuneration to the “original breeders”. CSIRO, in a 1993 media release speculated that introducing Boran and Tuli breeds could lift the production of the Australian herds by up to 30%. However, this was an ex-ante statement reflecting rather high-flying expectations, while reliable ex-post assessments of the eventual value added are not available. Another point of discussion is that the embryo and semen transfer technology has created both new advantages and disadvantages for the various stakeholders. Mpofu (2002) argued that the importing nations collect and process information concerning the African breeds and reinvest in their genetic improvement. They thus have a comparative advantage over the original breeders. Mushita (2005, pers. comm.) is concerned that useful genes of the Tuli might become patented without involving the local communities who developed the breed. He regards research on the Tuli by the private sector as a tool to gain monopolistic control over the breed. Moyo (2005, pers. comm.) seconds this by stating that Zimbabwe Tuli breeders had exciting years during the 1990s, selling their Tuli genetics onto the world market and that the interest was so great that Zimbabwe alone was not able to meet the demand. Farmers in Zimbabwe would believe that Australia who joined the race to export Tuli to America had an advantage in the negotiations because of their disease control situation and advances in technology.

444 CONCLUSIONS

6 CONCLUSIONS The lessons learned from the history and development of the Boran and Tuli cattle breeds, permit the following conclusions to be drawn: i. Improving the Boran and Tuli cattle breeds in their African environment is an example of successful within-breed selection and dissemination, maintaining a strong resistance to environmental stress. It is also an example for the comparative advantage of locally adapted breeds over imported exotic breeds. ii. The genetic material from the Boran and Tuli breeds – despite high expectations - does not seem to have significantly contributed to upgrading the international beef industry, although the actual impact cannot be quantified with the available data. However, the use of these breeds is an example for international gene transfers and the existing demand of industrial beef producers for new genetic material, which has particular beef quality traits and traits of adaptation to harsh environmental conditions. iii. The trade in Boran and Tuli genetic material has raised awareness of the potential of germplasm from developing countries also for livestock industry sectors in developed countries, and underlines the advantage of conserving livestock biodiversity for the future. The example has drawn public attention back to the value of Boran and Tuli purebreds in Africa. It has also contributed to raise awareness for adaptive traits and for beef quality traits in cattle breeding. iv. Embryo transfer and artificial insemination were the prerequisite for international trade of Boran and Tuli genetic material in that these techniques have permitted to overcome the sanitary barriers between the African countries and Australia. They have also reduced the costs and burden of live animal exports. This was crucial for the initial dispersion of Boran and Tuli germplasm to the world market. v. Vertical integration of multiple stakeholders - research institutions, government agencies, beef industry and individual business people - has essentially contributed to the initial transfer of Boran and Tuli genetic material to industrial countries as well as to their further marketing, lobbying and evaluation. Vertical integration could likewise be applied to breed conservation initiatives, involving local livestock keepers, breeders’ societies, universities, governments and development agencies. vi. The transfers of Boran and Tuli genetic material are characterised by free movements with governmental impact limited to livestock sanitary / veterinary regulations. Most transfers are between commercial stakeholders, and freely agreed benefit sharing can be assumed. Geographically the main streams were between Australia and America, and South Africa as well as within Africa. The countries of origin of the Boran and Tuli breeds are not the major players in this, however limited, business.

CONCLUSIONS 445

vii. The issues on whether or not a prior informed consent existed before the initial transfers of Boran and Tuli genetic material from Africa into Australia, and whether or not the call for additional compensatory payments to the original breeders is justified are controversial discussions. The main positions raised in the debate over the Boran and Tuli case are speculative and not based on sound scientific investigations. These issues are of an ethical nature, and have to be resolved politically. viii. The actual contribution of the Boran and Tuli to upgrading the Australian and American beef sector appears to be negligible. It did not nearly meet the high expectations which triggered the operation and the actual use of Boran and Tuli cattle in the Australian and American beef sectors is limited to singular cases. Prior informed consent should be more carefully looked for in future transactions to avoid a posteriori claims.

446 REFERENCES

7 REFERENCES

7.1 Personal communications de la Ray, M. 2005 : Personal communication. Getahun, T. and Ahmed, A. 2005: Personal communication.

Gibson, J. 2005: Personal communication.

Maynard, G. 2005 : Personal communication.

Mesfin, T. 2004: Personal communication.

Moyo, S. 2005 : Personal communication.

Mushita, A. 2005 : Personal communication.

Vercoe, J. 2004 : Personal communication.

Zander, K. 2005 : Personal communication.

7.2 Internet Agfact, 2004: http://www.agric.nsw.gov.au/reader/beef-breeds/a2339.htm, visited 30.11.2004.

All.Africa.com 2005: News item at All Africa com. http://allafrica.com, visited 05.01.2005.

ATSIC, 2004: http://www.atsic.gov.au/Programs/Industry_Strategies/rural_industry_strategy/Part_o ne.asp, visited 30.11.2004.

BCBS 2005: Boran Cattle Breeders Society of Kenya website at http://www.borankenya.org/research.htm, visited 15.01.2005.

Country World (2005): The rural newspaper of Texas. Internet media release 17.01.2002 at http://www.countryworldnews.com/Editorial/SCTX/2002/SC0117_Tuli.htm#top, visited 15.01.2005.

DAGRIS 2005: Domestic Animal Genetic Resources Information System (DAGRIS), Eds. J.E.O. Rege, W. Ayalew and E. Getahun, International Livestock Research Institute, Addis Ababa, Ethiopia, http://dagris.ilri.cgiar.org , visited 20.01.2005.

Daily Nation 1998: Beef Research Centre named 'most responsive'. Horizon. Internet media release 03.12.1998 at http://www.nationaudio.com/News/DailyNation/1998/031298/Features/XX9.html, visited 01.02.2005.

REFERENCES 447

Dallas Business Journal (2004): Internet media release at http://www.dallas.bizjournals.com/dallas/stories/2004/07/12/story4.htm, visited 01.02.2005.

Embryoplus, 2005: http://www.embryoplus.com/index.html, visited 03.01.2005.

FAO DAD-IS 2005: FAO Domestic Animals Database Information Service website at http://dad.fao.org/en/Home.htm, visited 15.01.2005.

FAO STAT, 2005: FAO Statistical Databases website at http://faostat.fao.org/faostat/, visited 24.01.2005.

Hammack 2003: Beef Cattle Browsing website at http://stephenville.tamu.edu/~shammack/newsletter/072597.htm#one, visited 08.11.2004.

Hill, 1993: African breed shows promise for carcass traits: http://agnews.tamu.edu/dailynews/stories/ANSC/african.htm, visited 30.11.2004.

ILRI 2005: International Livestock Research Institute (ILRI) website at http://www.ilri.org, visited 04.01.2005.

LPP, League for pastoral peoples: Securing tomorrow’s food. Promoting the sustainable use of farm animal genetic resources: http://www.pastoralpeoples.org/securing.htm, visited 25.11.2004.

TCBS 2004: Tuli Cattle Breeders Society, South Africa website at http://www.tuli.co.za, visited 20.12.2004.

TDA 2004: Texas Department of Agriculture website at http://www.agr.state.tx.us, visited 17.12.2004.

Temere Pastoral 2005: Temere Pastoral website at http://www.tremerepastoral.com.au, visited 25.01.2005.

Warrington, B. G., J. W. Holloway, D. W. Forrest and R. D. Randel. 2004: Lifetime performance of F1 tropically adapted cattle breeds x Angus in semi arid rangeland. Breeding and Genetics. Beef cattle research in Texas publication. http://animalscience.tamu.edu, visited 10.11.2004.

UniZim 2005: University of Zimbabwe website at http://www.uz.ac.zw/agriculture, visited 24.01.2005.

Zulu, H. 2003: Africa's Stolen Biodiversity. AfriDAD at http://asp.african- institute.org/debate_show.asp?Subject=Africa%B4s+Stolen+Biodiversity, visited 08.12.2004.

448 REFERENCES

7.3 Literature Aklilu, Y., P. Irungu, and A. Reda. 2002: An audit of the livestock marketing status in Kenya, Ethiopia and Sudan. Community-based Animal Health and Participatory Epidemiology Unit, Pan African Programme for the Control of Epizootics (PACE), Organisation of African Unity/Interafrican Bureau for Animal Resources (AU- IBAR), Nairobi, Kenya.

Alemu, G.B., A. Misgina, and B.I. Shapiro. 1998: Animal traction and the potential for use of dairy cows in the Ethiopian Highlands. In. Technology transfer: Multi-purpose cows for milk, meat and traction in smallholder farming system. Proceedings of an expert consultation held at ILRI. E. Zerbini, B. I. Shapiro, and J. C. Chirgwin, eds. Addis Ababa, Ethiopia, 11-14 September 1995.

Ali, K. O. 1998: Indigenous cattle of Zanzibar: the need for conservation. AGRI 23: 13-20.

Allen, J. 2001: Country Report: Australia. In. Allen, J. and A. Na-Chiangmai, eds. Development strategies for genetic evaluation for beef production in developing countries. Proceedings of an international workshop held in Khon Kaen province, Thailand, 23-28 July 2001. ACIAR Proceedings 108: 13-22.

Animal Productivity Research Unit. 1981: Ten years of animal production and range research in Botswana. Ministry of Agriculture, Gaborone, Botswana.

Animal Productivity Research Team. 1969: Zambia. Technical Report, 1968-1969.

Ayalew W. and J. Rowlands. 2004. Design, execution and analysis of the livestock breed survey in Oromiya Regional State, Ethiopia. OADB (Oromiya Agricultural Development Bureau), Addis Ababa, Ethiopia, ILRI (International Livestock Research Institute), Nairobi, Kenya.

Arnason, T. and H. Kassa-Mersha. 1987: Genetic parameters of growth of Ethiopian Boran cattle. Animal Production 44: 201-208.

Barrett, C., F. Chabari, D. L. Coppock, and P. Little. 2002: Livestock Pricing in the Northern Kenyan Rangelands. Mimeo, Cornell University, USA.

Behnke, R. H. and N. Abel. 1996: Revisited: The overstocking controversy in semi-arid Africa. World Animal Review 87: 4-27.

Bekure, S., P. N. de Leew, B. E. Grandin, and P. J. H. Neate. 1991: Massai herding: An analysis of the livestock production system of Masai pastoralists in Eastern Kajiado district. ILCA Systems Study 4, Addis Abeba, Ethiopia.

Bock, D. 1971: Untersuchungen über Einflüsse von Umwelt und Erbanlage auf das Absatzgewicht von Fleischrindern auf Ranchbetrieben in Kenia, Ph.D. thesis, University of Göttingen, Germany, 87 pp.

Briskin, J. 2002: Tuli cattle capturing wide interest. Central Texas Edition. The rural newspaper of Texas.

REFERENCES 449

Burrow, H. M., S. S. Moore, D. J. Johnston, B. Barendse and B. M. Bindon. 2001: Quantitative and molecular influences on properties of beef: a review. Australian Journal of Experimental Agriculture 41: 893-919.

Campher, J. 2003: Limitations and constraints in international trade in animal genetic material: a businessman's perspective. In. Generating Benefits through Sustainable Use and Conservation of Farm-Animal Genetic Resources in the SADC region: A workshop to explore the Legal and Regulatory Framework for Farm Animal Genetic Resources. Workshop held in Maputo Mozambique, 20-24 May 2003.

Commey, P. 2003: The new scramble for Africa. New African. December 2003 edition.

Coppock, D. L. 1994: The Borana plateau of southern Ethiopia: Synthesis of pastoral research, development and change, 1980-91. Addis Ababa, Ethiopia.

Cossins, N. J. 1985: The productivity of pastoral systems. Bulletin 21: 10-15.

Creek, M. J. 1977: The Kenya feedlot project. In. Animal breeding: selected articles from the World Animal Review. FAO Animal Production and Health Papers 1. FAO, Rome, Italy.

Cundiff, L. V., Gregory, K. E., Wheeler, S. D., Shackelford, S. D., Koohmaraie, M., Freetly, H. C., and Lunstra, D. D. 1996: Preliminary results from cycle V of the cattle germplasm evaluation program at the Roman L. Hruska U.S meat animal research centre. Germplasm Evaluation Program,Progress Report No. 15, USDA, Clay Center, NE, USA. de Leeuw, P.N. 1995: Maasai-Borana pastoralism (in preparation).

Demeke, S., W. C. Neser and S. J. Schoeman. 2004: Estimates of genetic parameters for boran, frisian, and crosses of frisian and jersey with the boran cattle in the tropical highlands of Ethiopia: milk production traits and cow weight. Journal of Animal Breeding and Genetics 120(4): 245-257.

Demeke, M. 1999: Agricultural technology, economic viability and poverty alleviation in Ethiopia. A paper presented at the Agricultural Transformation Policy Workshop, Nairobi, Kenya.

Dolan, R. B. 1997: The Orma Boran - a trypanotolerant East African breed. World Animal Review 89. 2 pp.

Dolan, R. B., H. Alushula, L. Munga, C. Mwendia, G. Okech, P. D. Sayer, P. G. W. Stevenson, R. L. Baker and M. Magadi. 1994: The Orma Boran - ten years of field observations. In Rowlands, G.J. and A.J. Teale, eds. Towards increased use of trypanotolerance: current research and future directions. ILRAD/ILCA. Nairobi. 71- 79.

Epstein, H. and I. L. Manson. 1971: The Origin of the Domestic Animals of Africa, Volume II. Africana Publishing Corporation, New York, USA.

450 REFERENCES

Fajersson, P., J. W. Holloway, R. D. Randel, J. Padilla, and C. R. Long. 1996: Use of newly aquired, tropically adapted germplasm in integrated production systems to improve beef cattle trade between Mexico and the USA. The British Veterinary Journal Online.

Felius, M. 1995: Cattle Breeds: An Encyclopedia. Misset uitgeverij, Doetinchem, NL.

Frisch, J.E. 1989: More meat in the Northern sandwich. Agricultural Science 3: 71-77.

Frisch, J. E. and C. J. O'Neill. 1998: Comparative evaluation of beef cattle breeds of African, European and Indian origins. 2. Resistance to cattle ticks and gastrointestinal nematodes. Animal Science 67(1): 39-48.

Frisch, J. E., R. Drinkwater, B. Harrison and S. Johnson. 1997: Classification of the southern African sanga and East Africa shorthorned zebu. Animal Genetics 28(2): 77-83.

Gande, C. 2002: Cattle breeder bows out. Daily News. Zimbabwe situation 15.05.2002.

Gazzola, C., C. J. O'Neill and J. E. Frisch. 1999: Comparative evaluation of beef cattle breeds of African, European and Indian origins. Animal Science 69: 135-142.

Goodwin, C.P.D. 1976: The Tuli – Rhodesia's masterpiece. Modern Farming 4: 9-13.

Gregory, K. E., J. C. M. Trail, H. J. S. Marples and J. Kakonge. 1985: Characterisation of breeds of Bos indicus and Bos taurus cattle for maternal and individual traits. Journal of Animal Science 60: 1165-1174.

Gura, S. and League for Pastoral Peoples. 2003: Losing livestock, losing livelihoods. Grain Seedling. GRAIN's quarterly magazine. January 2003.

Haile-Mariam, M., B. Malmfors and J. Philipsson. 1998: Boran - indigenous African cattle with potential. Currents 17/18: 38-43.

Haile-Mariam, M. and H. Kassa-Mersha. 1994: Genetic and environmental effects on age at first calving and calving interval of naturally bred Boran(zebu) cows in Ethiopia. Animal Production 58: 329-334.

Hanotte, O., C. L. Tawah, D. G. Bradley, M. A. Okomo, Y. Verjee and J. Ochieng. 2000: Geographic distribution and frequency of taurine Bos taurus and an indicine Bos indicus Y specific allele amongst sub-saharan African cattle breeds. Molecular Ecology 9(4): 387-396.

Hanotte, O., D. G. Bradley, J. W. Ochieng, Y. Verjee, E. W. Hill and J. E. O. Rege. 2002: African pastoralism: Genetic imprints of origins and migrations. Science 296(5566): 336-339.

Harvey, K. 1987: The history and development of the three indigenous cattle breeds of Zimbabwe. In Grant, J. ed. The indigenous cattle breeds of Zimbabwe, Proceedings of a workshop held at Matopos Research Station, 26-27 February 1987, Research and Specialist Services, Harare, Zimbabwe.

REFERENCES 451

Herring, A. D., J. O. Sanders, R. E. Knutson and D. K. Lunt. 1996: Evaluation of F1 calves sired by Brahman, Boran, and Tuli bulls for birth, growth, size, and carcass characteristics. Journal of Animal Science 74(5): 955-964.

Hetzel, D. J. S. 1988: Comparative productivity of the Brahman and some indigenous sanga and Bos indicus breeds of East and Southern Africa. Animal Breeding Abstracts 56(4): 243-255.

Homann, S. (2004). Indigenous knowledge of Borana pastoralists in natural resource management: a case study from southern Ethiopia. Cuvillier Verlag, Göttingen, Germany (PhD Thesis), 234 pp. http://geb.uni-giessen.de/geb/volltexte/2005/2112. Hunter, R. A. and B. D. Siebert. 1985: Utilization of low-quality roughage by Bos taurus and Bos indicus cattle. 2. The effect of rumen-degradable nitrogen and sulphur on voluntary food intake and rumen characteristics. British Journal of Nutrition 53: 649- 656.

Hunter, R. A. and B. D. Siebert. 1985: Utilization of low quality roughage by Bos taurus and Bos indicus cattle. 1. Rumen digestion. British Journal of Nutrition 53: 637-648.

Huqqa, G. 1999: The 37th gumi Gaayo assembly. Gada: The Oromo traditional, economic and socio-political system. The Norwegian Church Aid, Addis Abeba, Ethiopia.

Irungu, P. 2000: Cattle keeping practices of the Orma people. Results of a household survey in Tana River District, Kenya. KETRI-ILRI Collaborative Study. KETRI, Muguga. Kenya.

Köhler-Rollefson, I. 2000. Implementing the Convention on Biodiversity with respect to animal diversity. In. Experiences in Farmer’s Biodiversity Management: Report on the International Workshop on Animal and Plant Genetic Resources in Agriculture at the Biosphere Reserve Schorfheide-Chorin, Germany, 16-18 May 2000. German NGO Forum on Environment & Development, Bonn. 55-63.

Loftus, R. and E. P. Cunningham. 2000: Molecular Genetic Analysis of African Zebuid Populations. In. Blench, R. M. and K. C. MacDonald, eds. The Origins and Development of African Livestock: Archaeology, Genetics, Linguistics and Ethnography. London, UK. 251.

Lunstra, D. D. and L. V. Cundif. 2003: Growth and pubertal development in Brahman-, Boran-, Tuli-, Belgian Blue-, Hereford-, and Angus-sired F1 bulls. Journal of Animal Science 81: 1414-1426.

Marshall, F. 2000: The Origins and Development of African Livestock: Archaeology, Genetics, Linguistics and Ethnography. M. Blench and K. C. MacDonald, eds. Univ. College London Press, London, UK.

Mason I.L. and Maule J.P. 1960: The Indigenous Livestock of Eastern and Northern Africa. CAB, Farnham Royal, UK.

Mason, I. L. 1984: Evolution of domesticated animals. Longman, London, UK.

452 REFERENCES

Meyn, K. 1967: Weidemastversuche mit ostafrikanischen Zeburindern und Herefords, durchgeführt auf der Forschungsstation Muguga im Hochland Kenias, Ph.D. thesis, University of Göttingen, Germany, 183 pp.

Meyn, K. 1970: Beef Production in East Africa, Ifo – Institut für Wirtschaftsforschung, Afrika-Studienstelle, München, Germany. 227 pp.

Mhlanga, F. N., Khombe, C. T., and Makuza, S. M. 1999: Indigenous livestock genotypes of Zimbabwe. Departement of animal Science University of Zimbabwe, Mount Pleasant, Harare, 138 pp.

Moris, J. 1999: Under three flags: the policy environment for pastoralists in Ethiopia and Kenya. Unpublished SR/GL-CRSP Pastoral Risk Management Project Technical Report No. 04/99.

Moyo, S. 2003: Legal frameworks relevant to the management of farm animal genetic resources in the SADC region. In. Generating Benefits through Sustainable Use and Conservation of Farm-Animal Genetic Resources in the SADC region: A workshop to explore the Legal and Regulatory Framework for Farm Animal Genetic Resources. Workshop held in Maputo, Mozambique, May 20-24 2003.

Mpofu, N. 2002: The multiplication of Africa's indigenous cattle breeds internationally: the story of the Tuli and Boran breeds. Animal Genetic Traning, CD-Rom.

Nakimbugwe, H. 2003: The animal breeding act of Uganda. In. Generating Benefits through Sustainable Use and Conservation of Farm-Animal Genetic Resources in the SADC region: A workshop to explore the Legal and Regulatory Framework for Farm Animal Genetic Resources. May 20-24 2003; Workshop held in Maputo; Mozambique.

Nsoso, S. J. and T. G. Morake. 1999: A critical look at the use of exotic bulls in traditional farming in Botswana. South African Journal of Animal Science 29(2): 100-104.

Nwosu, C. C, F. N. Akhionbare and I. E. Iboh. 1983: Characterisation of cattle in Nigeria - body measurements. Tropische Landwirtschaft und Veterinärmedizin 23(1): 89-97.

Oba, G. 1998: Assessment of indigenous range management knowledge of the Booran pastoralists of southern Ethiopia. Borana Lowland Pastoral Development Programme (BLPDP/GTZ), Negelle, Ethiopia.

Oliver, J. 1983: Beef cattle in Zimbabwe. Zimbabwe. Journal of Agricultural Research 21: 1- 17.

Payne, W. J. A. and T. R. Wilson. 1999: An introduction to animal husbandry in the tropics. 5th ed. Blackwell Publishing Ltd, Iowa State University Press, U.S.

Pratt, D. J. and M. D. Gwynne. 1977: Rangeland management and ecology in East Africa. Hodder and Stoughton, London, UK.

Pretorius, L. 2004: Beste tuli's nou in SA. Landbouweekblad, South Africa.

REFERENCES 453

Pretorius, L. 2002: SA Boran kry Kenia se beste gene. Landbouweekblad, South Africa.

Pretorius, L. 2001: Groot taak vir SA Tuli telers. Landbouweekblad, South Africa.

RAFI 2000: Biopiracy - RAFI's Sixth Annual Update. RAFI Comunique.

Ralph, W. 1992: Australia's tropical cattle - the African Connection. Rural Research: a CSIRO quarterly.

Reda, A. 2001: Survey on the Borana cattle "breed" status to initiate community based conservation practices. Borana Lowlands Pastoral Development Programme (BLPDP/GTZ); Oromia Regional Bureau for Agricultural Development, Negelle, Ethiopia.

Rege J.E.O. 1998: The contribution of indigenous livestock to the economies of sub-Saharan African countries. In. CTA (Technical Centre for Agricultural and Rural Cooperation ACP-EU), Livestock development policies in eastern and southern Africa. Proceedings of a seminar organised by CTA, OAU/IBAR and the Ministry of Agriculture and Cooperatives, Swaziland, held at Mbabane, Swaziland, 28 July- 1 August 1997. CTA Wageningen, The Netherlands.

Rege, J. E. O., M. Kahi, M. A. Okomo, J. Mwacharo, and O. Hanotte. 2001: Zebu cattle of Kenya: Uses, performance, farmer preferences, measures of genetic diversity and options for improved use. Animal Genetic Resources Research 1. ILRI (International Livestock Research Institute), Nairobi, Kenya. 103 pp.

Rowlands, J. 1995: Field research in Keny on genetics of resistance to trypanosomias in East African cattle. Livestock Research in Development 1: 4-5.

Sacker, G. D., J. C. M. Trail and I. L. Fisher. 1971: Crossbreeding beef cattle in Western Uganda. Animal Production 13: 181-184.

Sandford, S. and Y. Habtu. 2000: Emergency response, interventions in pastoral areas of Ethiopia. Department for International Development (DFID), London, UK.

Senyatso, E. K. and B. S. Masilo. 1992: Animal Genetic Resources in Botswana. Department of Agricultural Research, Botswana.

Shapiro, B.I., E. Zerbini, and G. Takele. 1994: The returns to investment in the dual use of crossbred cows for milk production and draft in the Ethiopian Highlands. Proceedings of the First Conference of the Animal Traction Network of Ethiopia, Addis Ababa, Ethiopia, February 1994.

Shongolo, A. 1995: The Gumi Gaayo assembly of the Boran: a traditional legislative organ in the modern world. Zeitschrift für Ethnologie 119: 26-52.

Sisay, A. and Ebro, A. 1988: Growth performance of Boran and their Simmental crossbred calves at Adami Tulu. Proceedings of the Annual Conference of Ethiopian Society of Animal Production, Addis Ababa, Ethiopia, 14-15 May 1995.

454 REFERENCES

Tawonezvi, H. P. R. 1984: Crossbreeding range beef cattle for weaner production in Zimbabwe. Ph.D. thesis, University of Zimbabwe.

Tawonezvi, H. P. R., H. K. Ward, J. C. M. Trail and D. Light. 1988: Evaluation of beef breeds for rangeland weaner production in Zimbabwe. I. Productivity of purebred cows. Animal Production 47: 351-360.

Taylor, R. 1994: New breeds for northern beef industry. Rural Research 162: 10-11.

Tegegne, A., A. Geleto, P. O. Osuji, T. Kassa and R. Franceschini. 1994: Influence of dietary supplementation and partial suckling on body weight and on lactation and reproductive performance of primiparous Boran (Bos indicus) cows in Ethiopia. Journal of Agricultural Science 123: 267-273.

Teka, T., A. Azeze, and A. Gebremariam. 1999: Cross-border livestock trade and food security in the Southern and South-eastern Ethiopia borderlands. OSSREA development research report series ed. Organisation for Social Science Research in Eastern and Southern Africa (OSSREA), Addis Abeba, Ethiopia.

Thorpe, W., D. K. R. Cruickshank and Thompson R. 1980: Genetic and environmental influence on beef cattle production in Zambia. 1. Factors affecting weaner production from Angoni, Barotse and Boran dams. Animal Production 30: 217-234.

Thorpe, W., D. K. R. Cruickshank and Thompson R. 1981: Genetic and environmental influence on beef cattle production in Zambia. 4. Weaner production from purebred and reciprocally crossbred dams. Animal Production 33: 165-177.

Tonn, R. 1974: Die Eignung des Boranrindes und seiner Kreuzungen mit europäischen Leistungsrassen zur Fleischerzeugung an extensiven tropischen Standorten, untersucht in Ranchbetrieben des Rift Valleys in Kenia, Ph.D. thesis, University of Göttingen, Germany, 237 pp.

Trail, J. C. M. 1984: In. Baker, F. H. and M. E. Miller, eds. Beef Cattle Science handbook, Vol. 20. Westview Press Inc., Boulder, Colorado, 297-300.

Trail, J. C. M, Buck, N. G., and Light, D. 1980: Future beef cattle breeding systems- Botswana. Consultancy report covering cattle breeding research assistance to Government of Botswana by the International Livestock Centre for Africa (ILCA), October 1980.

Vercoe, J. 1992: Animal Adaptation - the development of the beef industry in Northern Australia. Agricultural Science, special issue on international agriculture: 50-56.

Wanyama, J. 2003: Generating Benefits through Sustainable Use and Conservation of Farm Animal Genetic Resources in the SADC Region : In. Generating Benefits through Sustainable Use and Conservation of Farm-Animal Genetic Resources in the SADC region: A workshop to explore the Legal and Regulatory Framework for Farm Animal Genetic Resources. Workshop held in Maputo, Mozambique, 20-24 May 2003.

REFERENCES 455

Zerbini, E., A. Gebre Wold, and B. I. Shapiro. 1998: Development of cow traction technologies for adoption in the East African Highlands’. In. Zerbini, E., B. I. Shapiro, and J. C. Chirgwin, eds. Technology transfer: Multi-purpose cows for milk, meat and traction in smallholder farming system. Proceedings of an expert consultation at ILRI. Addis Ababa, Ethiopia, 11-14 September 1995.

456 CONTACT ADDRESSES

8 CONTACT ADDRESSES

Homann, Sabine Maritz, Jacobus Hendrik International Crops Research Institute for the Institute of Animal Production in the Tropics Semi-Arid Tropics (ICRISAT), and Subtropics, University of Hohenheim, POB 776, Bulawayo 70593 Stuttgart ZIMBABWE GERMANY Phone: +26 3 11623967 Phone: +49 711 4593172 Fax: +26 3 838253/8307 Fax: +49 711 4593290 Email: [email protected] Email: [email protected]

Hülsebusch, Christian G. Meyn, Klaus Centre for Agriculture in the Tropics and ADT Projekt GmbH, Adenauerallee 174, Subtropics, University of Hohenheim, 53113 Bonn 70593 Stuttgart GERMANY GERMANY Phone: +49 228 9144730 Phone: +49 711 4593742 Fax: +49 228 9144731593 Fax: +49 711 4593315 Email: [email protected] Email: [email protected] Valle Zárate, Anne Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart GERMANY Phone: +49 711 4593170 Fax: +49 711 4593290 Email: [email protected] Ahmed, Abdi Ayalew, Workneh Pastoral Forum Ethiopia (PFE), Panos ILRI-BMZ Project Manager POB 1570 code 1110 or 1152, Addis Ababa Animal Genetic Resources Group ETHIOPIA International Livestock Research Institute Phone: +251 11 666363/59 (ILRI) Fax: +251 11 666361 POB 5689, Addis Ababa Email: [email protected] ETHIOPIA Phone: +251 11 6463215 Fax: +251 11 6461252/ 6464645 Email: [email protected]

CONTACT ADDRESSES 457

de la Ray, Morné Getahun, Tezera EMBRYO PLUS Executive Director, Pastoral Forum Ethiopia Phone: +27 12 250 2359 (PFE), Panos Fax: +27 12 250 2299 POB 1570 code 1110 or 1152, Addis Abeba POB 2644, BRITS, 0250, North West, ETHIOPIA SOUTH AFRICA Phone: +251 11 666363/59 Email: [email protected] Fax: +251 11 666361 http://www.embryoplus.com Email: [email protected] Email: [email protected] Maynard, Geoff Mesfin, Tafesse Mount Eugene Belmont Reds & 5 Star FARM-Africa Ethiopia Senepols POB 5746 "Mount Eugene" Addis Ababa POB1, Jambin 4702 Ethiopia Queensland, AUSTRALIA Phone : +251 11 155 26 84 Phone: +61 7 4996 5240 Fax: +251 11 155 21 43 Fax: +61 7 4996 5316 Email: [email protected] Email: [email protected] Moyo, Siboniso Mushita, Andrew Dept. of Livestock Production Commutech Bevan Building POB 7232, Harare, ZIMBABWE 18 Borrowdale Road Phone: +263 4 589 256 Harare, ZIMBABWE Fax: +263 4 589 390 Email: [email protected] Email: [email protected] Vercoe, John Wanyama, Jacob Dr John E Vercoe AM FTSE ITDG-EA (Intermediate Technology 7 Ryan Street, Zilzie Q 4710 Development Group East Africa) AUSTRALIA POB 39493, Postal Code: 00623 Phone/Fax: +61 7 49387486 Nairobi, KENYA Email: [email protected] Phone: +254 2 2713540 Fax: +254 2 2710083 Email: [email protected] Zander, Kerstin Center for Development Research (ZEF) Walter-Flex-Straße 3 53113 Bonn GERMANY Phone: +49 228 731852 Email: [email protected]

458

GENE FLOW IN ANIMAL GENETIC RESOURCES. A STUDY ON STATUS, IMPACT AND TRENDS

Editors: Anne Valle Zárate, Katinka Musavaya and Cornelia Schäfer

Institute of Animal Production in the Tropics and Subtropics, University of Hohenheim, Germany

ANNEX 9.9

UNIVERSITY OF HOHENHEIM Institute of Animal Production in the Tropics and Subtropics

Impact of the use of exotic compared to local pig breeds on socio-economic development and biodiversity in Vietnam Le Thi Thanh Huyen, Regina Roessler, Ute Lemke, Anne Valle Zárate

VERLAG ULRICH E. GRAUER ‚ Beuren ‚ Stuttgart ‚ 2005

Le Thi Thanh Huyen, Regina Roessler, Ute Lemke, Anne Valle Zárate: Impact of the use of exotic compared to local pig breeds on socio-economic development and biodiversity in Vietnam

VERLAG GRAUER, Beuren, Stuttgart, 2005.

ISBN 3-86186-496-7

© 2005 Institut für Tierproduktion in den Tropen und Subtropen Universität Hohenheim (480a), 70593 Stuttgart, Deutschland E-mail: [email protected]

All rights reserved.

Printed in Germany. Druck: F. u. T. Müllerbader GmbH Forststr. 18, 70794 Filderstadt, Deutschland

VERLAG ULRICH E. GRAUER Linsenhofer Str. 44, 72660 Beuren, Germany Tel. +49 (0)7025 842140, Fax +49 (0)7025 842499 Internet: http://www.grauer.de/, E-Mail: [email protected]

This case study is an independent part of the gene flow study implemented by the Institute of Animal Production in the Tropics and Subtropics of the University of Hohenheim. The Fed- eral Ministry for Economic Cooperation and Development (BMZ) and the German Technical Cooperation (GTZ) acted as commissioner and project executing agency. The Food and Agriculture Organisation (FAO) acted as a support agency. An advisory panel composed of international scientists, representatives of donor and development agencies, the private sector and NGOs closely accompanied the study. The co-funding by the Sonderforschungsbereich 564 (SFB 564) promoted by the Deutsche Forschungsgemeinschaft (DFG), is gratefully acknowledged, as is the support of all spon- sors.

ACKNOWLEDGEMENTS In Vietnam, we wish to thank Dr. Le Thi Thuy and Dr. Nguyen Van Dong of the National Institute of Animal Husbandry (NIAH) in Hanoi and the Department of Agriculture and Ru- ral Development in Son La for supporting the successful completion of the case study by providing valuable inputs. We gratefully acknowledge contributions from all before mentioned institutions and persons to this study.

TABLE OF CONTENTS 465

TABLE OF CONTENTS

List of tables 467

List of figures 467

Abbreviations 467

Executive summary 469

1 Introduction 471 2 Formation and distribution of local breeds 472 2.1 Delta areas, northern Vietnam 473 2.1.1 I pig 473 2.1.2 Mong Cai 474 2.1.3 Lang Hong 475 2.1.4 Thai Binh 475 2.1.5 Tap Na 475 2.2 Central and northern mountains of Vietnam 476 2.2.1 Meo 476 2.2.2 Muong Khuong 476 2.2.3 Co 477 2.2.4 Soc 477 2.2.5 Tuy Hoa 477

3 Formation and distribution of composite breeds 478 3.1 Thuoc Nhieu 478 3.2 Ba Xuyen 478 3.3 Trang Phu Khanh 479 3.4 DBI-81 and BSI-81 479 3.5 Improved Mong Cai 480

4 Introduction of high performance breeds and crossbreds 482

5 The pig breeding system and its stakeholders in Vietnam 488

6 Suitability of different breeds for different environmental conditions 490 6.1 Conditions of smallholder pig production in Vietnam 490 6.1.1 Feeding systems and management 490 6.1.2 Pig housing 491 466 TABLE OF CONTENTS

6.1.3 Diseases 491 6.2 Productive and reproductive performance of different genotypes in different production systems 491

7 Impact of imports on biodiversity 496

8 Conclusions 499

9 References 501

10 Contact addresses 507 LIST OF TABLES 467

LIST OF TABLES Table 1: Local pig breeds in Vietnam 472 Table 2: Chronology of creation of Ba Xuyen and Thuoc Nhieu pigs 479 Table 3: Import of exotic pig breeds to Vietnam 483 Table 4: Reproductive performance of different pig genotypes in Vietnam 493 Table 5: Live weight gain of different pig genotypes in Vietnam 494

LIST OF FIGURES Figure 1: Distribution of major local breeds in Northern Vietnam 472 Figure 2: Distribution of major local breeds in Central Vietnam 473 Figure 3: Mong Cai sow 475 Figure 4: Ban sows 476 Figure 5: Pig population and liveweight of pigs in Vietnam from 1975 to 1999 486

ABBREVIATIONS ACIAR Australian Centre for International Agricultural Research AI Artificial insemination IFAD International Fund for Agricultural Development LR Landrace pig breed LW Large White pig LW gain Liveweight gain MC Mong Cai pig breed NIAH National Institute of Animal Husbandry, Vietnam N Vietnam North Vietnam PIC Pig Improvement Company SIDA Swedish International Development Cooperation Agency SU Soviet Union S Vietnam South Vietnam VND Vietnamese Dong VSF Vétérinaires sans Frontières 468 ABBREVIATIONS

EXECUTIVE SUMMARY 469

EXECUTIVE SUMMARY This case study focuses on the formation and distribution of the main indigenous pig breeds and crossbreds in Vietnam, the introduction of high performance breeds and their impact on biodiversity, and the suitability of different breeds for different environmental conditions. Vietnam owns a wide variety of local pig breeds across different regions of the country. The Lang Hong, Mong Cai and I breed are the product of a long deliberate breeding history, whereas other breeds, e.g. the Meo, Co, or Soc, were not systematically bred. In particular the I and later the Mong Cai were strongly promoted in Vietnam to replace lower yielding local breeds. In South Vietnam, the Thuoc Nhieu, Ba Xuyen and Phu Khanh composite breeds developed from crossbreeding local with exotic pigs. The DBI-81 and BSI-81 were developed in North Vietnam from crossbreeding I sows with exotic boars, but did not be- come widespread in national pig production. Only the Mong Cai has become common, being now the major local sow line in Vietnam. Exotic pigs, including Large White, Landrace, Du- roc and Berkshire, have been introduced to Vietnam from American and European countries since before the 1920s. Major driving forces were the French Colonial Rulers (before 1954), American forces (before 1973), the socialist government (since 1954), Vietnamese and for- eign commercial companies (before 1954 and after 1986), and developmental projects (after 1986). Gene flow now and recently is mainly a net inflow of exotic pigs. Current develop- ment and poverty alleviation projects at village level usually promote exotics, and only occa- sionally improved Vietnamese breeds. Information on pig gene flow to and within Vietnam is limited, due to the restricted information policy of both international breeding companies and Vietnamese official sources, but also due to the decentralised nature of pig breed import and distribution. At present, exotic and crossbred pigs dominate, while local pigs make up only 26% of the national pig herd, mostly in uplands, rural and remote areas. The decentralised structure of the Vietnamese breeding system, the less developed central coordination and the common use of AI have all supported the spread of exotic pigs in Vietnam, especially at the small- holder level which makes up 80 to 95% of Vietnamese pig production. Smallholder pig production includes different intensity levels. In contrast with large-scale commercial pig production, they can be characterised as low-input systems. Local pigs yield lower reproductive and growth performances. Performance data in literature are rarely com- parable, as local breeds were usually investigated in low-input extensive-farming conditions, while exotic pigs or crossbreds are often tested under improved conditions or on station. Mong Cai sows under smallholder conditions yield higher reproductive performances than exotic or crossbred pigs, implying better reproductive performance potential of local breeds. Additionally, favourable adaptation traits (regarding environmental/climatic factors, low in- put production conditions, and susceptibility to disease) and general robustness are described for local pig breeds, together with favourable meat quality traits. Other, less favourable traits of local pig breeds include a high fat content and low lean meat ratio, a low growth rate, and, apart from the Mong Cai, a low fertility, rendering them less suitable to respond to higher inputs, unless their special quality traits are rewarded by the consumer. The influx of exotic breeds had a strong impact on local pig populations. Today, 10 of 14 local pig breeds are in vulnerable or critical state or face extinction, and all of them show declining populations. NIAH is the main Vietnamese institution conducting conservation 470 EXECUTIVE SUMMARY programs, but only for a limited number of pig breeds. The long-term sustainability of those programs is questionable. The significant genetic distinctions both between Vietnamese breeds and between Vietnam- ese and European breeds have been shown. Local breeds are a source of promising alleles, which might be significant for future genetic improvement and of unpredictable economic value. Local pig breeds are a significant component of the Vietnamese and worldwide biodiversity, and are still important for resource poor farmers in Vietnam, who depend on them to ensure their livelihoods. The dominance of high yielding exotic breeds will increase in intensified production systems. Local breeds will only contribute to worldwide biodiversity if their com- petitiveness to exotics is proved for production systems under development and/or if favour- able adaptation traits are proved and the controlling alleles identified. Investigations are un- der way to define local pig breeds, characterise them, and compare their performances under standardised conditions.

INTRODUCTION 471

1 INTRODUCTION This case study focuses on the formation and distribution of the main indigenous pig breeds and crossbreds in Vietnam, the introduction of high performance breeds and its impact on biodiversity, and the suitability of different breeds for different environmental conditions. Livestock forms 25 percent of the agricultural output value in Vietnam and is almost entirely in the hands of small farmers. Traditional farming systems integrate crops, fish and livestock, mainly pigs and poultry (Ly, 1996). Industrial livestock production on state-run farms and large-scale private farms is developing, but is still only a minor part (Thong, 1996). For a long time, animal production in Vietnam has been based on local genotypes (Lemke et al., 2000), which are well adapted to local climates and available inputs. However, their pro- ductivity under improved conditions is lower than that of exotic breeds (Ly, 1996), under low-input conditions it might be comparable or even higher than that of exotics, but both are rarely tested together under same low-input conditions. Thus, local breeds have been re- placed or crossbred with imported high-yielding breeds to increase performance in recent decades, leading to a severe decrease in the number of indigenous breeds, which are an im- portant part of Vietnam’s biodiversity (Lemke et al., 2000). In 2002, the Vietnamese pig population was 23 million head. Indigenous pigs accounted for 26% of the total, located mainly in uplands, rural and remote areas. Local breeds can utilize farm-grown feeds and by-products, survive and produce in low-input systems, and withstand hardships. In rural and mountainous areas, local pigs have a multitude of functions, including consumption, capital storage, use in weddings, funerals, religious celebrations, as gifts and as suppliers of manure (Country Report of Vietnam, 2003). That report mentions 14 indigenous Vietnamese pig breeds, five of them in vulnerable state, two in critical state, and three facing extinction. Disappearance of local breeds might put smallholders’ food security and econo- mies at risk, or might be just a consequence of a better suitability of exotic breeds to meet farmers demands. Thuy (2004) showed that Vietnamese indigenous breeds were genetically distant to European breeds, had a higher number of alleles per gene locus, and were genetically more heteroge- neous than European breeds. The large genetic distance between selected breeds can be ex- ploited in crossbreeding, benefiting from heterosis and combination effects for performance and adaptation traits. This case study describes the formation and distribution of local breeds (chapter 2) and of composite breeds (chapter 3), and the introduction and spread of exotic high performance breeds to Vietnam (chapter 4). Chapter 5 describes the pig breeding system and its stake- holders in Vietnam. Chapter 6 discusses the suitability of selected pig breeds under different environmental conditions met in Vietnam. Chapter 7 discusses the impact of importing ex- otic breeds on biodiversity. Conclusions are drawn from the presented results.

472 FORMATION AND DISTRIBUTION OF LOCAL BREEDS

2 FORMATION AND DISTRIBUTION OF LOCAL BREEDS According to ecological and economic conditions, Vietnam is divided into seven agro- ecological zones: Northern Mountainous and Middle Highlands, Red River Delta, Northern Central Coast, Southern Central Coast, Central Highlands, North-East of Southland and Me- kong Delta (Ly, 1996). There are diverse local pig breeds in the different regions of Vietnam (Table 1).

Table 1: Local pig breeds in Vietnam Uplands, Central and Northern Vietnam Southern Vietnam Northern Vietnam I Meo Ba Tri Mong Cai Co Muong Khuong Soc Lang Hong Tuy Hoa Meo (Ban, H’mong) Tau Pha Thai Binh Nghia Binh Tap Na Mini pig/ Quang Tri Tong Con Ha Bac Son Vi

Source: compiled from: Molenat and Thong, 1991; Ly, 1999; Ly, 1999; FAO/DAD-IS, 2004; Thuy, 2004) Figure 1 and Figure 2 depict the distribution of the major local pig breeds in Vietnam.

Figure 1: Distribution of major local breeds in Northern Vietnam

(Upper case letters mark distribution area of pig breeds: A = I pig, B = Mong Cai pig, C = Lang Hong pig FORMATION AND DISTRIBUTION OF LOCAL BREEDS 473

Figure 2: Distribution of major local breeds in Central Vietnam

Upper case letters mark distribution area of pig breeds: A = Meo/Ban pig, B = Muong Khuong pig, C = Co pig, D = Soc pig According to Lemke et al. (2000), the indigenous breeds Lang Hong, Mong Cai and I can be characterised as improved, higher-yielding breeds resulting from a longer systematic breed- ing period, compared to the Meo and Co indigenous breeds which have not been deliberately improved. Other local pig breeds include the Tong Con at the Chinese border, the white Nghia Binh pig (central Vietnam) and the Ha Bac, a small breed predominantly kept in the central region of northern Vietnam (Molenat and Thong, 1991). The Tau Pha breed predomi- nates in smallholder scavenging systems in the mountainous regions of central Vietnam. In the central highlands, a black and white pig resembling the Mong Cai has been observed but whose breed has not yet been identified (Hot, 1982). Thuy (2004) mentions additionally the Son Vi pig (Phu Tho province) and the Mini pig of Quang Tri province.

2.1 Delta areas, northern Vietnam

2.1.1 I pig The I pig originates from Nam Dinh province (nowadays Nam Ha), Red River Delta. It is a very small, black pig with potbelly and swayed back (Ly, 1999). Two sub-species can be distinguished (Khanh and Hien, 1963), namely the I-mo (or pure I pig) and the I-pha. The I-mo is smaller than the I-pha with short legs and trunk, a big head with short snout bent upward and small, upright ears. In the extremely wrinkled face, the eyes seem to be closed. Most I-mo sows have 10 teats. 474 FORMATION AND DISTRIBUTION OF LOCAL BREEDS

The I-pha is a cross of I-mo with various (unknown) local breeds, but is nowadays consid- ered a distinct breed. It has longer legs, trunk and snout than the I-mo. The snout tapers. The potbelly is less pronounced and the face is only slightly wrinkled. The ears are bigger and stand horizontally. I-pha sows have 10 to 12 teats. The I pig is characterized by early maturity, high fecundity and adaptation to a frequently flooded, muddy environment and poor, roughage-based diet. I pigs are said to be resistant to parasites (Tang and Cuong, 1994). In general, they are considered very robust. The I pig has a high fat and low lean meat proportion (Molenat and Thong, 1991). Before 1970, the I pig was most common in North Vietnam. The province of origin (Nam Dinh, nowadays Nam Ha) had good access via roads and rivers to the other delta provinces, thus supporting the introduction of the I pig to Ha Nam, Ha Tay, Hung Yen, Hanoi, Vinh Phuc, Hai Duong, Thai Binh, Quang Ninh, Ninh Binh, Hai Phong and Thanh Hoa (Tang and Cuong, 1994). 1970-1989: The promotion and distribution of the higher-yielding Mong Cai (“Mong Cai- isation”) lead to a parallel reduction of the I pig population. In 1989, only 463 sows were counted in the original breeding region (Hoang Hoa district). In this time, the National Insti- tute of Animal Husbandry Hanoi (NIAH) started conservation measures to protect the breed from extinction (Lemke et al., 2000). In 1989, 22 I-mo pigs (2 boars, 20 sows) were selected from the remaining I pig population of two villages to build a nucleus (Tang and Cuong, 1994). Due to internal problems, that conservation project was dissolved in 1994, and 20 sows and two boars from the nucleus were transferred to a new nucleus in Hoang Hoa. Since 1989 (start of conservation), the nu- cleus has slightly expanded; in 2000, the nucleus consisted of 36 sows and six boars (Lemke et al., 2000). No data exist on the current population size. There remain only few communes in Thanh Hoa province where I pigs are kept in small numbers outside the conservation pro- gram (Ly, 1999).

2.1.2 Mong Cai The Mong Cai, today the main breed in North and central Vietnam, originated from North East Vietnam (Duyet and Duong, 1996), from the Red River delta, coastal provinces Hai Phong and Thai Binh (Thien et al., 1996). It has a small to medium body size, and small, upright ears. Head and body are black, with a white band running from one side of the ab- domen over the shoulder to the other side of the abdomen, making a black saddle over the middle of the swayed back (Figure 3). The Mong Cai is characterised by high prolificacy. It is adapted to poor quality feed, and is in general robust. Disease resistance has been reported but not scientifically confirmed. FORMATION AND DISTRIBUTION OF LOCAL BREEDS 475

Figure 3: Mong Cai sow

Picture taken at smallholder households of ethnic Black Thai in Son La province, North West Viet- nam (picture: Lemke)

Two types of Mong Cai have been described, a small-frame and a large-frame. The large- frame type is said to have a higher lean meat rate, a bigger litter size and a higher growth rate. However, piglet mortality in the large-frame type is also higher (Hai et al., 1979). The pure Mong Cai pig with a small body size originated from the sea shore region of Quang Ninh province (formerly Hai Ninh province), bordering Quang Dong province (China). The Tong Cuu pig of Quang Dong province resembles the Mong Cai pig (Doanh, 1994). 1960-1975: Since the 1960s, the Mong Cai has spread throughout the northern delta areas. After 1975: Mong Cai pigs were introduced to Central and South Vietnam (Ly, 1999). The Mong Cai was frequently used to improve local breeds with lower performances. Thus, the number of pure Mong Cai pigs gradually declined, while the number of Mong Cai crossbreds (with other local or exotic breeds) increased (Doanh, 1994).

2.1.3 Lang Hong The Lang Hong pig resembles the Mong Cai in appearance and performance. However, there has not yet been an attempt to assess any relation between the breeds. Compared to the latter, the Lang Hong has a shorter trunk, less pronounced potbelly and swayed back, a shorter snout, and smaller, upright ears. The forehead bears a white triangular mark. Lang Hong pigs are mostly raised in Bac Ninh and Bac Giang provinces (North East Viet- nam). At present, Lang Hong pigs are mostly crossed with Mong Cai (Ly, 1999).

2.1.4 Thai Binh The Thai Binh breed originates from the Red River delta. It is a small pig of white colour with black marks, has a swayed back and a pronounced potbelly (Molenat and Thong, 1991).

2.1.5 Tap Na The Tap Na has been found in Cao Bang province and neighbouring mountainous provinces (North East Vietnam), kept under low-input conditions. It is a black pig with six white marks (forehead, four feet, tip of the tail). The Tap Na shows some external features resembling the Mong Cai, but has a black belly and no white band at the shoulders like the latter one. The Tap Na has a straight head of medium size with stooping ears, the snout is not wrinkled. It has long legs like the Muong Khuong or Meo. Its back is straight and the belly does not touch the ground. It has 6 to 10 teats. The Tap Na is well adapted to the ecological conditions where it is kept. It is said to be resis- tant to certain diseases, and to have a tasty meat. Daily weight gain and lean meat percentage are low and the feed conversion rate is high. 476 FORMATION AND DISTRIBUTION OF LOCAL BREEDS

The number of Tap Na pigs is decreasing and it is at high risk of extinction. Nowadays, Tap Na pigs in villages near main roads are crossbreds, and pure Tap Na are difficult to find (Duc et al., 2004).

2.2 Central and northern mountains of Vietnam

2.2.1 Meo The Meo pig originated from the high mountainous areas of Truong Son. Meo pigs are mainly kept by ethnic Thai and H’mong people in North and North West Vietnam, who call their pigs by different names, e.g. Ban, Dan, Meo (with different diction) and H’mong. It remains to be investigated, whether these are different breeds, eco-types of the same breed, or one breed under different names (e.g. Hoa, forthcoming). In this study, the term Meo is used. Meo pigs have also spread to the mountainous areas of Nghe An province (Central Vietnam), an area populated by Thai farmers. The Meo is well adapted to the local ecologies and socio-economies of the H’mong people (Ly, 1999). Meo pigs resemble wild boars. They have no potbelly and a straight back (Figure 4). Thai farmers describe three local varieties differing in size, appearance of white marks (snout, tip of the tail, legs), reproductive performance and growth rate (Lemke et al., 2000). For Meo pigs of the H’mong, six phenotypic groups have been described, which might repre- sent sub- or eco-types (Huyen, 2004).

Figure 4: Ban sows

Pictures taken at smallholder households of ethnic Black Thai in Son La province, North West Vietnam (Lemke, 2002)

Meo sows are said to have less favourable mothering abilities. They reach maturity late (8 to 9 months). Under the husbandry conditions of H’mong farmers, litters are with 6 to 7 piglets small; the farrowing interval is high. About 60 to 70% of piglets survive until weaning. H’mong boars are sexually mature at four to five months age (To and Duc, 1967). Meo boars have been mated with sows of other local breeds for commercial purposes (Ly, 1999).

2.2.2 Muong Khuong The Muong Khuong pig closely resembles the Meo pig. It is kept by H’mong farmers in the mountainous areas of North Vietnam, particularly Lao Cai province. The Muong Khuong is a large pig and copes well with being kept as a scavenger. Due to similarities between Muong Khuong and Meo concerning performance, appearance, adaptation traits, and area of keep- ing, Vietnamese scientists have tried to identify whether they belong to the same breed (Ly, FORMATION AND DISTRIBUTION OF LOCAL BREEDS 477

1999). The development plans of Lao Cai province for the period 2001 to 2005 include a plan for conserving the Muong Khuong, mainly to preserve it as a sow line for crossbreeding (Thu, 2004).

2.2.3 Co The Co pig is found in the Central Highlands of Vietnam. Its very small body size and low performance is most likely a result of inbreeding (Hot, 1982). Before 1960, Co pigs were common in the central provinces Nghe An, Ha Tinh and Binh Tri Thien, but were strongly reduced in numbers due to the spread of the Mong Cai to central provinces (Ly, 1999).

2.2.4 Soc The Soc pig also originates from the Central Highlands. It is kept by the ethnic groups of E De, Gia Rai, Ba Na, and Mo Nong, in the provinces Lam Dong, Dak Lak, Gia Lai and Kon Tum. Soc pigs have a small body and resemble wild pigs. They are kept as scavengers, with- out supplemental feeding. Due to the introduction of higher-yielding pigs, the population of Soc pigs in the Central Highlands has strongly decreased (Ly, 1999).

2.2.5 Tuy Hoa The Tuy Hoa originates from the Song Ba river delta. Tuy Hoa pigs have little hair and a white colour and are quite sensitive to insolation. They show a good growth performance under good feeding conditions (Molenat et Thong, 1991). Under smallholder conditions, Tuy Hoa pigs show a great variation in performance. 478 FORMATION AND DISTRIBUTION OF COMPOSITE BREEDS

3 FORMATION AND DISTRIBUTION OF COMPOSITE BREEDS In Vietnam, a number of composite breeds are known. However, breeding history and ge- netic make-up in some cases rather suggest that breed standardisation has not yet been ac- complished. The Ba Xuyen and Thuoc Nhieu breeds (Mekong delta) and the Trang Phu Khanh (Central Vietnam), Thuoc Nhieu and Ba Xuyen breeds were formed from spontane- ous natural mating between local sows and exotic boars in the 1920s (Doanh et al., 1985). They have widely spread since due to good adaptation to the local climate, high prolificacy, good mothering abilities and a high growth rate (Tjällden, 1999). More recent attempts of Vietnamese scientists to create synthetic breeds from I sows and exotic boars date back to the 1980s. Further, the improved Mong Cai with exotic influence in its genetic make-up can be distinguished from the original Mong Cai.

3.1 Thuoc Nhieu The Thuoc Nhieu is a medium-sized pig (120 to 160 kg adult liveweight) of white colour. It has small upright ears, a medium-length snout, short legs and a long body with slightly swayed back. Growth rate and reproductive performance are moderate (litter size: 8 to 10 piglets/litter), the lean meat ratio is low and the fat percentage high (Molenat and Thong, 1991). The Thuoc Nhieu can cope with feedstuffs of low nutritional value and harsh keeping conditions (Tjällden, 1999). In 1900, the Chinese imported pigs to the coastal areas around the Mekong delta. Vietnamese local black sows (probably Co) were mated with those Chinese Hainan boars. The resulting crossbreds (F1, called Tau Pha, black-and-white pigs) were mated with French Craonnais boars (imported by the French, now extinct). The resulting F2 crossbred (called Bo Xu) was then continuously crossed with imported Yorkshire Large White and Yorkshire Middle White, creating the Thuoc Nhieu pig, which stabilized after some decades (FAO, 1999). The Thuoc Nhieu was popular in the provinces Tien Giang and Long An and the vicinity of Ho Chi Minh City, and later spread to the provinces Vinh Long, Long An, Dong Nai, Binh Thuan, Can Tho, and Soc Trang (Mekong River Delta) (Ly, 1999). The Mekong Delta is one of the most important pork-producing areas of Vietnam. To fulfil national policies and meet the increasing market demand, scientists of the Institute of Agri- cultural Science of South Vietnam carried out selection programs mainly between 1981 and 1989. In this period, the Thuoc Nhieu’s growth and reproductive performance improved by over 10% compared to those in smallholder conditions, and the breed then stabilized com- pared to those kept under smallholder conditions. The improvements are due to both success- ful breeding practices and the higher input on station. Thuoc Nhieu pigs for breeding were selected from rural backyards in Tien Giang province based on body conformation, growth rate, litter performance (sows), semen quality and sired litter performance (boars). First- grade boars and sows of second grade and upwards were placed in state breeding farms of the province for inbreeding between lines for two or more generations with continuing selec- tion. The results showed increases in weaned litter weight and litter size. Selected sows also showed good prolificacy, surpassing Yorkshire sows reared in Vietnam (Thong et al., 1996).

3.2 Ba Xuyen Like the Thuoc Nhieu, the Ba Xuyen was created by mating Vietnamese native black sows (probably Co) with Chinese Hainan boars and mating the resulting F1 (Tau Pha) with French FORMATION AND DISTRIBUTION OF COMPOSITE BREEDS 479

Craonnais boars. Between 1932 and 1956, mating the F2 (Bo Xu) with American Berkshire boars yielded the Ba Xuyen, a black pig with white spots, appreciated by farmers (Hai, 1994). Ba Xuyen pigs were concentrated in Soc Trang province. At present, they are spo- radically raised in the provinces Vinh Long, Can Tho, Tien Giang, Kien Giang, An Giang, Long An, and Dong Thap (Mekong River Delta) (Ly, 1999). Table 2 summarises the creation of Ba Xuyen and Thuoc Nhieu pigs (Doanh et al., 1985).

Table 2: Chronology of creation of Ba Xuyen and Thuoc Nhieu pigs

Year Formula of crossing 1900 Chinese Hainan (B) x Vietnamese local black (S) ↓ 1920 French Craonnaise (B) x F1 (S) ↓ F2 (Bo Xu) 1932-1956 American Berkshire (B) x F2/Bo Xu (S) ↓ Ba Xuyen pig 1936-1956 Yorkshire Large White (B) x F2/Bo Xu (S) 1957 Yorkshire Middle White (B) x F2/Bo Xu (S) ↓ Thuoc Nhieu pig Abbr.: B = boar, S = sow

3.3 Trang Phu Khanh The Trang Phu Khanh pig originates from Phu Khanh province (today provinces Khanh Hoa and Phu Yen, South Central Coast) as a crossbred between Yorkshire and local pigs. It is still kept in the area, being a common sow line (Duyet and Duong, 1996). A long period of cross- breeding and selection led to the development of the Trang Phu Khanh, resembling in per- formance and appearance the Yorkshire. The Trang Phu Khanh has been recognized as Viet- namese pig breed since 1989. However, breeding efforts were not well coordinated, and Trang Phu Khanh pigs increasingly mixed with other pig breeds (Ly, 1999). Trang Phu Khanh sows have a good fertility and mothering abilities (Duyet and Duong, 1996).

3.4 DBI-81 and BSI-81 In 1981, scientists of the National Institute for Animal Husbandry (NIAH) created two new pig types: the white-coloured DBI-81 from mating Vietnamese I sows with Soviet Large White boars, and the black-coloured BSI-81 from mating I sows with Berkshire boars. They intended to create pigs with higher lean meat ratio than the local pigs and higher robustness than the exotic breeds. DBI-81 and BSI-81 were employed to generate crossbred offspring from local I and Mong Cai sows for fattening, acceptable to farmers and suitable for hus- bandry conditions in Vietnam (Doanh and Thong, 1985). NIAH selected DBI-81 and BSI-81 480 FORMATION AND DISTRIBUTION OF COMPOSITE BREEDS boars and supplied them to the artificial insemination network in northern and central prov- inces (Thuong, 1985). The genotypes were most widely spread in Hanoi and the provinces Ha Son Binh, Ha Nam Ninh, Ha Bac, and Thanh Hoa (North Vietnam) (NIAH, 1985).

3.5 Improved Mong Cai With comparatively favourable characteristics, particularly a high reproductive performance, the Mong Cai pig has been used to improve other local pig breeds with lower performances, especially in North and Central Vietnam. Mong Cai are mainly used as a sow line (Thien et al., 2002). Over time, the number of pure Mong Cai gradually declined, while the population of Mong Cai crossbreds (with local or exotic pigs) strongly increased (Doanh, 1994). Since 1959/60, many scientific studies have been conducted on the Mong Cai. Mong Cai under- went a breeding and selection process at the state breeding farms including Thanh To, Dong Trieu, and Tam Dao through activities such as breeding assessment, boar examination and creating the nucleus herds (Doanh et al., 1985). The following examples document the distri- bution process. Since 1975, Mong Cai were exported to Central Vietnam by state breeding farms (Dong Trieu, Tam Dao, Thach Ngoc). Through selection and adaptation in this centre, the Mong Cai has been shown to have a good mothering ability, large litters, a good milk production, a high number of litters per year, and a large body size (Duyet and Duong, 1996). In 1975 and based on government policies, 300 Mong Cai sows from the provinces Tam Dao, Quang Ninh and Hai Phong (North Vietnam/Red River Delta) were brought to a state breeding farm in Hue (Central Vietnam) to replace local pigs of low production. However, this farm col- lapsed in 1980, and a number of sows were given to local farmers. Between 1977 and 1987, pig breeding in Central Vietnam focussed on Mong Cai and Trang Phu Khanh. However, breeding efforts were limited to the government/state farms; there were no national policies concerning farmers. (Duyet and Duong, 1996). Since the 1990s, Mong Cai and (Yorkshire x Mong Cai) sows have been imported to Thanh Hoa province (North Central Coast), as part of the province’s developmental program, to increase the lean meat rate in the province’s pig herd (Luong and Gian, 1999). In Tuyen Quang province (North East Vietnam), Mong Cai sows have been introduced since 1992 (Minh, 2000; Dong and Tiep, 2002). The breeding program of Tuyen Quang province was an open nucleus breeding program carried out in two periods. From 1992 to 1994, 69 Mong Cai sows and two Mong Cai boars of associated farms were given to 69 smallholders in Son Duong district, which received additional funding. The associated farms ceased work in 1994, but from 1995 to 1998 the number of pure Mong Cai sows amongst the smallholders increased very quickly. The breeding program expanded to include many smallholders of other districts of Tuyen Quang province, reaching a breeding population of over 3 millions Mong Cai sows. This strong development was partly due to external support, as the program received funding e.g. from the IFAD project of the province (International Fund for Agricul- tural Development) and SIDA (Swedish International Development Cooperation Agency). Smallholders received credits from the project for keeping Mong Cai sows (1 mVND/sow/year, no interest rate). Further project activities included developing AI cen- tres to distribute Mong Cai semen in 6 towns and districts of Tuyen Quang province, to cre- ate a pure Mong Cai herd of high quality. FORMATION AND DISTRIBUTION OF COMPOSITE BREEDS 481

Government policy in 1994 no longer included maintaining Cornwall pigs, but confirmed the importance of Mong Cai pigs for the national breeding strategy due to its good reproductive performance and adaptation traits. However, male Mong Cai yield very low prices due to their inferior fattening performance, and market demand is only for female Mong Cai for breeding. Thus, keeping a purebred Mong Cai population is not economically desirable for farmers (Rodríguez et al., 1996). For example, in Thua Thien/Hue province, only one private farm keeps one Mong Cai boar, and the whole province as well as farmers from other prov- inces depend on this Mong Cai boar. The availability of semen is low (2 doses each 3 days, at a price of VND 8,000 = US$ 0.62/dose). Demand for Mong Cai in the provinces is consid- ered to be very high. 482 INTRODUCTION OF HIGH PERFORMANCE BREEDS AND CROSSBREDS

4 INTRODUCTION OF HIGH PERFORMANCE BREEDS AND CROSSBREDS The indigenous Vietnamese pig breeds are characterized by a low growth rate and high car- cass fat (Xuan et al., 1995). Exotic breeds like Yorkshire, Berkshire, Cornwall and Landrace have been introduced as part of government strategies and to satisfy the growing demand for large lean carcasses. However, they are less adapted to the local environment and husbandry conditions than the native breeds, and so yield performances lower than their genetic poten- tial (Singh et al., 1996). About 10% of the Vietnamese breeding herd consist of adapted ex- otic breeds, which have been kept a considerable time in Vietnam under local conditions, mainly in state farms and small-scale private farms, concentrated in South East Vietnam (Hai, 1996). From 1970 onwards, crossbreds of local sows and improved or exotic boars with hybrid vigour have been increasingly used (Quac et al., 1996). Table 3 summarises the import of exotic breeds to Vietnam. INTRODUCTION OF HIGH PERFORMANCE BREEDS AND CROSSBREDS 483

Table 3: Import of exotic pig breeds to Vietnam

Year Pig breeds Origin Vietnamese organization facili- tating the breed import 1920 Craonnais France Individual farms, S Vietnam 1932 Berkshire USA 1936 Yorkshire USA 1955 LW, Berkshire France Phat Ngan Animal Husbandry Corporation, S Vietnam 1957-1959 Yorkshire Japan 1964 LW, Berkshire SU NIAH, N Vietnam 1965-1966 Yorkshire, LW, Hampshire, USA Phat Ngan Animal Husbandry Chester White, Duroc, LR, Corporation, S Vietnam Poland China 1969 Yorkshire USA Phat Ngan Animal Husbandry Corporation, S Vietnam 1971 LW SU NIAH, N Vietnam Berkshire China Landrace Cuba 1976 Duroc S Vietnam Agricultural University No. I and III, N Vietnam 1977 Yorkshire, LR, Duroc Cuba NIAH, N Vietnam 1978 LW, Duroc Cuba NIAH, N Vietnam 1997 Yorkshire, LR, Pietrain England PIC, N Vietnam 2000 Duroc, LW USA NIAH, N Vietnam Abbr.: LR = Landrace, LW = Large White, NIAH = National Institute of Animal Husbandry, PIC = Pig Im- provement Company, SU = Soviet Union Source: compiled after: Doanh (1985); Thien et al. (1992); Hai (1996); Truc et al. ( 2003) Before 1954 (end of French colonisation), exotic breeds were imported by individual farms in South Vietnam. At that time, Ba Xuyen and Thuoc Nhieu pigs were developed (see above) mostly to meet the domestic market demand, but also partly for export. After 1954, when US forces occupied South Vietnam, exotic breeds were mainly imported from the US to meet the demand of the US Army. As a result, pig production developed strongly, particularly from 1964 to 1967 (1964 to 1973 American War). Exotic pigs were imported through Phat Ngan Animal Husbandry Corporation, South Vietnam, and were then bred both pure and crossed, creating the so-called Yorkshire Phat Ngan pig. From 1965 to 1974, Yorkshire Phat Ngan had a great impact on South Vietnamese pig production. Industrial pig production developed with large and medium scale farms especially around Saigon (today Ho Chi Minh City). Several factories produced commercial feedstuffs; Phat Ngan Animal Husbandry Corpora- tion supplied breeding animals, and a number of companies processed pork. After 1975 (end of the American war, reunification of Vietnam), Soviet Large White pigs were imported 484 INTRODUCTION OF HIGH PERFORMANCE BREEDS AND CROSSBREDS through NIAH (11 boars, 87 sows) and brought to Lam Dong province (North East South) (Doanh, 1985). Between 1975 and 1986, French Landrace pigs were introduced to South Vietnam through the France Hybrid Company. From 1964 to 1977, North Vietnamese state farms imported mainly exotic boars for mating with local sows. NIAH directly managed those imports. In 1976, Duroc pigs from the South were introduced to the North and raised in the research farms of Agricultural University No. I and III (Truc et al., 2003). Imported pigs were kept and tested in the breeding centres of NIAH and other state farms in Hanoi and Hai Hau under intensive conditions, in order to create parent stock to supply breeding animals and F1 crossbreds for farmers and commercial animal production (Doanh and Luan, 1985). Results showed that, in general, imported breeds could adapt to Vietnamese climatic and husbandry conditions. However, compared to pure parent stock in the countries of origin, the performance of animals kept in Vietnam declined by 20 to 30% due to lower nutritional levels and the tropical climate. For example, in im- ported Large White pigs the reproductive performance of sows declined. In contrast, the se- men quality of Large White boars raised in Vietnam nearly equalled that of Large Whites in the Soviet Union. Large White pigs born and raised in Vietnam developed an outer appear- ance with thinner hair and a lighter body conformation. Large White boars were used to cre- ate DBI-81 pigs (see above). Imported Berkshire pigs also adapted to local climatic and hus- bandry conditions at state farms. They were very robust, and especially insusceptible to mange. Berkshire boars were used to mate I and Mong Cai sows. However the reproductive performance declined in the Berkshire sows as well. Imported Duroc yielded lower perform- ances and had higher mortalities. Imported Yorkshire, Landrace and Duroc often developed respiratory or reproductive diseases during adaptation periods at state farms, but disease in- cidence has decreased over the years of raising exotic pigs in Vietnam (Doanh, 1985). In 1997, the British PIC (Pig Improvement Company) introduced 480 great-grandparent pigs to North Vietnam (Tam Diep farm, Ninh Binh province) including the lines L11 (Yorkshire), L06 (Landrace), L64 (Pietrain) and the composed lines L19 (Duroc/Yorkshire) and L95 (Landrace/Chinese Meishan). In July 2001, all these lines were transferred to Vietnam through NIAH. At present, Tam Diep farm produces grandparent stock for Vietnamese breeding farms, which in turn raise parent stock for smallholders, which then raise and fatten the end products (Truc et al., 2003). Another farm with 600 great-grandparent sows was es- tablished by PIC Ltd. Vietnam in Dong Giao (Ninh Binh province) before 2000. CP group supported Vietnam with the construction of a pig AI station for 50 exotic boars in Hung Yen province for Hung Yen and neighbouring provinces (Lich and Tuyen, 2001). From 1995 to 2001, the Australian Centre for International Agricultural Research (ACIAR) funded a project on “Breeding and feeding pigs in Australia and Vietnam” (AS2/1994/023), involving collaboration between the Institute of Agricultural Sciences of South Vietnam and the Queensland Department of Primary Industries. About 40 Australian Large White and Duroc pigs were brought to Vietnam and crossed with Vietnamese breeds to produce lean pigs with higher growth rate. A system of performance testing and selection was imple- mented. ACIAR and the Australian Agency for International Development (AusAID) pro- vided funding to support AI centres. In addition to the breeding-related activities, two ‘least- cost’ diets were formulated based on traditional and non-traditional ingredients (ACIAR, 2004). INTRODUCTION OF HIGH PERFORMANCE BREEDS AND CROSSBREDS 485

Parallel to developing crossbreds, and based on governmental policy, Vietnamese scientists started to introduce AI from 1958 onwards (Thien, 2002). Since the 1980s, the advantages of AI have been increasingly recognized. Crossbred feeder pigs increased in number, reaching 60% of the total herd (Thong, 1996). The Yorkshire was considered as a base breed in the development of two-way and three-way crossbreds. These crossbreds help to increase pork production and carcass quality and fulfil consumer demand for meat. The use of exotic pigs and crossbreds and of advanced management techniques was supported by the Vietnamese research and education system, through the agricultural extension service (training and tech- nical advice), and the mass media (radio, television). Networks of technical staff and dem- onstration farms have been established for appropriate feeding, rearing management and a supply of certified boars, plus AI services and disease management (Hai, 1996). Previously, traders and farmers used to sporadically introduce exotic pigs from South Viet- nam to the Central Highlands, including Large White, Yorkshire, Duroc and Landrace. Due to uncontrolled breeding and extensive management, exotic pigs intermingled with each other and with local pigs, leading to the great variety in the pig population found today. After 1975, exotic breeds (Cornwall, Large White, Yorkshire, Landrace) were imported to Central Vietnam (Duyet and Duong, 1996), and state breeding centres were established e.g. in Duc Trong (Large White) and Bao Loc and Buon Ma Thuot (Edelschwein). Among the breeds brought to the Central Highlands, the Cornwall seemed to adapt best to local conditions, but had a high carcass fat content, comparatively lower fertility and undesired colour, and thus did not get general approval for breeding. Large White, Yorkshire and Landrace were mainly used for AI service. The proportion of exotic blood in fatteners increased considerably. Crossbred fatteners now yield growth rates of 430 to 500g/day, a lean meat ratio of 50%, and have a low feed consumption. Large Whites have become more widespread than the other exotic breeds in the Central Highland (Hot, 1982). From 1981 to 1989, the Institute of Agricultural Science of South Vietnam carried out a pro- ject of selective breeding to improve Yorkshire pigs in Ho Chi Minh City and neighbouring provinces (Thong et al., 1996). The breeding was based on pigs from commercial breeding farms (216 sows, 23 boars) and family farms (1050 sows, 40 boars). Selected pigs had higher performances than the population average (litter weight at birth: + 16%, litter size at wean- ing: + 9%, litter weight at weaning: + 17%). Further crossbreeding experiments were con- ducted between Thuoc Nhieu sows and Yorkshire and Landrace boars; 419 crossbred litters were performance tested. The F1 crossbred offspring had higher performances than the pure parental pig breeds (Thong et al., 1996). The so-called Lean Meat Program was based on a governmental decision to increase, over a 10-year-period (1990 to 2000), the lean meat proportion in the Vietnamese pig herd. Exten- sive investigations between 1990 and 1994, e.g. at NIAH and Bac Thai experimental farm, proved that crossbreds between local and exotic pig breeds were well adapted to economic and ecological conditions in the Red River Delta. Particularly the three-way cross of (Land- race x (Large White x Mong Cai)) was chosen for widespread use in breeding programs in the Red River Delta Region in the following decade 2000 to 2010 (Thien et al., 1996). The northern mountainous provinces received a large number of new pigs, including exotic imports and Mong Cai pigs. For example, the People’s Committee Son La imported Russian Large White, Belgian Yorkshire, Hungarian Cornwall, American Duroc and Danish Land- race (source: Son La Department of Agricultural and Rural Development) through the state breeding centre and regional breeding centres of the province, agricultural extension centres 486 INTRODUCTION OF HIGH PERFORMANCE BREEDS AND CROSSBREDS of province and districts, national development and poverty alleviation projects and private organisations of the province. Imported pigs were kept in regions near towns with compara- tively good infrastructure and economic conditions. In 1994, 556 pigs (mainly Landrace, Yorkshire) were brought from South Vietnam to state farms in Thanh Hoa province (Huy et al, 1996). In several pilot projects, exotic pigs were introduced to smallholder households, e.g. in 1988, the Vietnam Institute of Agricultural Science and Technique introduced Yorkshire sows to smallholders in Hai Duong province (Red River Delta) (Duy et al., 2001). Selected house- holds were relatively well-off and located in regions with already strongly developed pig production. Contracts were made with the farmers to keep the exotic sows. The risk was shared between farmer and the research institute; no additional direct funding was involved. Interested non-project farmers were advised how to select, buy, keep and manage exotic sows. Results proved the adaptation abilities of Yorkshire to smallholder conditions; farmers yielded high outputs and an acceptable efficiency. In the following year, this model was ap- plied widely in the province. More breeds were introduced in 1995 and 1996. The Thuy Phuong pig research centre (NIAH) introduced pure Landrace and Yorkshire pigs to farmers in Red River Delta prov- inces (e.g. Ha Tay, Thai Binh) and investigated their performances under smallholder condi- tions. Smallholders tended to use extensive farming techniques even for the exotic pigs. After more than 40 years of research a number of crossbreds have been identified that com- bine the favourable genetic characteristics of local and exotic breeds, in both state farms and smallholder farms and in different eco-systems. In particular, government research between 1990 and 1995 produced the crossbreds LR x (LW x MC), LW x (LW x MC), LR x (LR x (LW x MC)) and LR x (LR x (LR x MC)) for commercial production. In 2000, the cross- breds (Duroc x (LW x LR) and Duroc x (LR x YR)) were introduced into pig production. The results of crossbreeding programs led to changes in the breeding strategies for both sows and fatteners (Thien, 2002).

Figure 5: Pig population and liveweight of pigs in Vietnam from 1975 to 1999

Pig population Pig liveweight (in 1000 heads) (in 1000 tons) 20000 1400

18000 1200 16000

14000 1000

12000 800

10000

600 8000

6000 400

4000 Pig population 200 2000 Pig liveweight

0 0 1975 1980 1985 1990 1995 Source: General Statistical Office (2000) INTRODUCTION OF HIGH PERFORMANCE BREEDS AND CROSSBREDS 487

The growth of the Vietnamese pig population increased markedly since 1992, probably re- flecting the economic changes after 1986 (Figure 5). The plot depicting the pig liveweight runs almost parallel to the plot depicting the popula- tion, indicating that a higher production output is mainly due to an increasing pig population. The rise in the pig liveweight and convergence of the two plots reflect the higher percentage of exotic and crossbred animals in the total population, especially in the last decade (see be- low, chapter 7). 488 THE PIG BREEDING SYSTEM AND ITS STAKEHOLDERS IN VIETNAM

5 THE PIG BREEDING SYSTEM AND ITS STAKEHOLDERS IN VIETNAM The national pig breeding system of Vietnam has mainly been implemented at state farms (under government administration) and provincial farms (administered by the People’s Committee of the province) (ASPS, 2002). Before 1995, Vietnam had 53 state breeding farms; almost all of them keeping Yorkshire, Landrace, Duroc and Mong Cai. The exotic pigs had been imported from France, Belgium and other European countries, Japan, Thailand and the US. These state breeding farms supply piglets for fattening, and produce boars for both natural mating and AI (Lich, 1996). The majority of breeding centres are involved in several levels of the breeding pyramid. Commercialisation has required many breeding cen- tres (Decision 68/1998/QD-TTg, 1998) to change from breed development and dispersal of genetics to production and sale of commercial stock for fattening. The emphasis on short term commercial gain is at the expense of a long-term national vision of livestock improve- ment. Only a small number of breeding centres and research institutes keep local breeds for crossbreeding and conservation. The Vietnamese pig breeding is not centrally coordinated. Neither genetic improvement nor breed replacement and conservation are uniform across regions (ASPS, 2002). A certain amount of the breeding centres’ output reaches farmers via the mass organisations (mainly Women’s Union, Farmers’ Association) and national devel- opment projects. For example, the national extension service runs a Breed Distribution Pro- ject, with the aim to supply one Large White boar to each village. Almost all northern and central provinces have AI stations, each holding up to 20 breeding boars; individual AI stations keep as many as 30 to 100 boars. Insemination services in North Vietnam are more developed than in the South. Boars for AI are mainly of Yorkshire, Land- race and Duroc genotype and produce in total over 1 billion doses per year. However, many AI stations lack equipment for semen collection and processing, resulting in poor semen quality (Lich, 1996). In 2000, Vietnam had 265 district AI stations, 138 of them in the North (including independent AI stations). There were 10 state breeding farms with a total of 2,000 grandparent sows including indigenous, exotic and hybrid genotypes, 2 boar testing stations with a test capacity of 300 boars/year, and 10 provincial breeding farms keeping 1,500 breed- ing sows (Lich and Tuyen, 2001). As an example, AI is most common in the Hue-Thua Thien province: farmers obtain semen from the AI centre in Hue city, and inseminate sows themselves. However the whole prov- ince has only 3 AI centres and this limits the availability of semen (Rodríguez et al., 1996). The state farm in Hue city has 8 boars (7 Large White, 1 Mong Cai). In total, 40 doses Large White/day (VND 6,000/dose = US$ 0.46/dose) are distributed to four technicians in 4 dis- tricts, who distribute them further. Veterinarians and farmers conduct the insemination at village level. In Quang Phuoc village, a provincial breeding farm was rented to a farmer, who manages the farm himself, providing 25 doses/day (Large White). Another private farmer keeps a Yorkshire boar, supplying semen at VND 4,000/dose (US$ 0.31). Renting AI sta- tions to private persons and private boar ownership was also seen in Son La province and is probably widespread. Since the late 1980s with the start of the open door policy in Vietnam, foreign governmental and non-governmental organisations have played a role in introducing and distributing higher-yielding pig genotypes in a large number of both big and small projects. Examples include the activities of ACIAR and AusAID (see above), the IFAD Country Program Viet- nam (IFAD International Fund for Agricultural Development; Tuyen Quang province) or THE PIG BREEDING SYSTEM AND ITS STAKEHOLDERS IN VIETNAM 489

SEDEC (Socio-Economic Development Center for Costal Areas) in cooperation with Kon- rad-Adenauer-Stiftung (Binh Thuan province, 2001). With project activities starting in 2001, Vétérinaires sans Frontières have promoted Mong Cai sows in Phu Tho province (North East Vietnam), now extending their activities to other North Vietnamese provinces. In the field of international companies and breeding organisations, active importers include PIC, CP Thailand, the Dutch TOPIGS (export via TOPIGS international in Canada and the US), the United Kingdom Pig Breeding Association (Nhien, 2004) and the Danish DanBred. It is assumed that a great number of other international traders are actively introducing exotic pigs to Vietnam, but information on those activities is scarce. 490 SUITABILITY OF DIFFERENT BREEDS FOR DIFFERENT ENVIRONMENTAL CONDITIONS

6 SUITABILITY OF DIFFERENT BREEDS FOR DIFFERENT ENVIRONMENTAL CONDITIONS Today, Vietnam owns a number of indigenous pig breeds. Depending on the various cli- mates, ecological and socio-economic conditions in the distribution areas, there is consider- able variation in performance from one breed to another (Ly and Duyet, 2000). Each region has its own special conditions and over the generations different breeds have adapted to their unique environments as people create domestication and breeding processes to suit their own specific needs. As a result, breeds have been developed that produce even under extreme conditions, are robust, show low susceptibility or even resistance to disease, and can survive on limited nutritional resources (Tjällden, 1999). However, ecological and economic condi- tions are changing, and animals adapted to production systems that no longer exist will be replaced by those better suited to the new, prevailing systems. Hai and Nguyen (1997) described three production systems in Vietnam: state run farms (4 to 5% of total pig production), private commercial farms (15%) and smallholders (80%). In 1998, around 95% of sows in Vietnam were kept in extensive households, while less than 5% were kept under intensive conditions (Pig International, 1998). Lemke et al. (2002) and Valle Zárate et al. (2003) indicated that pig production in Son La province (North West Vietnam) showed different levels of production intensity. Lemke et al. (2002) described a semi-intensive system used in mountain valleys and near towns, where pig production is driven by generating income from the sale of pork. Introduced Mong Cai and Mong Cai crossbreds dominate. In the extensive systems found at hillsides and further away from town, pig production is driven by availability of resources. Local pig breeds pre- vail (Meo/Ban), serving various functions (consumption, income generation, social func- tions).

6.1 Conditions of smallholder pig production in Vietnam Major constraints to livestock production in Vietnam concern animal nutrition, health, live- stock genetic potential (FAO, 1999) and marketing (Lich and Tuyen, 2001; Vang, 2002).

6.1.1 Feeding systems and management Smallholder pig production is mainly based on the utilisation of farm-produced feedstuffs and agricultural by-products, characterised by a high fibre content and low protein and en- ergy contents (Loc et al., 1996). Feedstuffs for pigs include rice bran, broken rice, maize, vegetables, agricultural by-products like soybean cake, fish meal, salted fish waste, and commercial concentrates. Particular for lactating sows, farmers provide protein-rich feed supplements (Thong, 1996), however, at a limited amount (Rodríguez et al., 1996). Fishes and shrimps (fresh or dried) are used sporadically as protein supplements (Peters, 1998). Households with larger pig herds and engaging in activities like wine production, tofu proc- essing or grain milling, can use by-products as pig feed; which has been observed in both lowlands and highlands (Tung, 1999). It has even been observed by the authors of this study that families started distilling and selling rice wine in order to make fermented rice as pig feed. At smallholder households in Son La, pig feeding was based on maize, rice bran and cassava. Comparing Thai villages near town, Thai villages in intermediate location, and re- mote H’mong villages, the use of commercial feed decreased with increasing remoteness, SUITABILITY OF DIFFERENT BREEDS FOR DIFFERENT ENVIRONMENTAL 491 CONDITIONS and the use of garden and forest vegetables increased. In Thai villages, between 40 and 90% of pigs were fed on purchased feedstuff with seasonal variations, while in remote H’mong villages it was only 4% (Huyen, 2004; Lemke et al., forthcoming). In general, feeding depends on the crop season and the family’s condition (Rodríguez et al., 1996). Monetary investment into feeding by smallholders is low (Tung, 1999; Ly, 2000).

6.1.2 Pig housing Pig housing at smallholder level is simple (Rodríguez et al., 1996). According to Astroem (2000), prevailing systems in rural areas are free range systems or simple pens, both with a minimum of inputs. Pig housing at smallholders’ shows a great variation: H’mong and Thai farmers keep pigs as scavengers, occasionally confining them in paddocks; in wooden or bamboo-made pens, of- ten stilted; in pens with concrete floor, wooden/bamboo-made fence and canvas/tile/asbestos roof; or brick-built stables (Lemke et al., 2000; Huyen, 2004). In the Central Highlands, most common are stables with a packed clay-floor, less frequently with a partially concreted floor and a partially packed-clay floor. Some pens are surrounded by a fence (often a ‘live’ fence of cassava or bamboo) to restrict the area in which piglets can scavenge (Rodríguez et al., 1996). Pig housing in Tuyen Quang province is on a closed concreted floor. Compartments are constructed either from bamboo or concrete; the roof is made from leaves, sometimes from tiles (Bosma et al., 2003). Smallholders in Ha Tay province (Red River Delta) were observed by the authors of this study to construct massive brick stables with concrete floor, tile roof, and compartmented by brick walls.

6.1.3 Diseases Animal diseases are a risk to livestock development, with a relatively high mortality rate causing considerable losses of the GDP (NIAH, 2003). Low vaccination coverage has been identified as a major problem (Dung, 2002). Vaccination coverage for Vietnam in general in the 1990s was given with 40 to 50% (Thuy, 1999), and for North Vietnam with 25% (Lich and Tuyen, 2001). Among prevalent diseases, Foot-and-Mouth disease FMD and Classical Swine Fever are endemic (source: http://www.oie.int). Resistance of local pig breeds against diseases and endoparasites is often mentioned by Vietnamese sources (e.g. the I pig is said to be resistant against FMD), but not scientifically proven. High mortalities do not support such perceived disease resistance, but probably result from unimproved, input-extensive manage- ment (Lemke et al., 2000).

6.2 Productive and reproductive performance of different genotypes in different pro- duction systems In Vietnam, pig production is based on two groups of breeds, the native and the exotic breeds. In Vietnam, native breeds are usually smaller than imported breeds. On the other hand, favourable characteristics including adaptation to climate and low-input production systems, robustness and lower susceptibility against diseases have been described (while not always been scientifically proven). Native pigs are able to thrive on poor quality feeds, and are productive under conditions where imported breeds would not (or would even not sur- vive). Compared with exotics, local breeds can cope with lower amounts and qualities of inputs (e.g. feed, veterinary inputs) and less intensive care. It is for those favourable charac- 492 SUITABILITY OF DIFFERENT BREEDS FOR DIFFERENT ENVIRONMENTAL CONDITIONS teristics that farmers appreciate local breeds (Thong et al., 1996). Exotic breeds require more intensive feeding and management in order to yield performances according to their higher genetic potential (Quac et al., 1996). In their countries of origin they are highly productive, but are difficult to raise under unfavourable conditions in Vietnam. It is estimated that exotic pigs in Vietnam realise only 60 to 70% of their performance potential, leading to low eco- nomic efficiency (Thien et al., 1996). In addition to reduced performance, mortalities of ex- otic pigs under tropical conditions are higher than in their native countries. However, it may be that local pigs are only kept where there are no alternatives: if available, farmers keep exotic pigs because of their higher performances and reputation and in order to benefit from subsidies attached to those exotics. Some local breeds have a reasonable reproductive performance and produce good quality meat. Disadvantages include a small body size and a low production output per time unit (Ly, 1993). Table 4 and Table 5 give an overview of reproductive and growth performance for different pig genotypes. As different genotypes were kept in different production systems, the performances should not be directly compared. In addition, for the growth performance data, age classes were not distinguished. The improved Mong Cai has a higher production output but requires a higher input. The local Meo yields a lower output but needs less input (Lemke et al., 2002). Crossbreds of (Large White x Mong Cai) genotype in Central Vietnam had significantly higher daily gains, when fed on protein supplements (Loc et al., 1996). Mong Cai pigs are prolific, having large litters, and are robust. They seem to be less suscep- tible towards diseases. Disadvantages include a slow growth (liveweight at 10 months 60 to 68 kg), high feed conversion ratio (4.5 to 5.0 kg concentrate/kg LW gain) and low lean-meat rate of 32 to 34% (Thien et al., 1996). To benefit from the favourable reproduction and adap- tation characteristics of Mong Cai, while improving the undesirable fattening and carcass traits, Mong Cai sows are commonly mated with exotic boars like Large White, Danish Landrace, Cornwall, Hampshire or Duroc (Astroem, 2000). Meo pigs have a reputation of superior adaptation and robustness and can, like I and Co pigs, cope with a fibre-rich diet. Their meat quality is considered excellent. Anh and Dung (1994) give a lean meat ratio of 50 to 60%, higher than in other local breeds (the ages of pigs with this performance is missing). Meo sows farrow 5 to 15 piglets/litter, in 1.0 to 1.4 litters/year. Long farrowing intervals are due to high weaning ages (> 2 months), and to the fact that of- ten male piglets sire the next litter after becoming sexually mature. The pre-weaning mortal- ity is 20% (Anh and Dung, 1994; Thuy, 1999). According to To and Duc (1967), Meo pigs in Nghe An province (Central Vietnam) have growth rates of 4 to 8 kg/month (age 2 to 4 months) and of 12 to 15 kg/month (age 6 to 8 months); and Meo fatteners reach 250 kg liveweight at 12 to 18 months age; corresponding to growth rates of 133 to 266 g/day (2 to 4 months), 400 to 500 g/day (6 to 8 months), and 694 to 463 g/day for the total lifetime. It is assumed that those results are overestimated, especially in comparison with other data in Table 5. The favourable adaptation traits of the I pig have been mentioned above. I pigs reach sexual maturity early with 3 to 4 months age, have a high reproductive performance, good mother- ing abilities and a remarkable longevity (Tang and Cuong, 1994; Ly, 1999). In contrast, Lemke et al. (2000) found a moderate litter size and especially high piglet mortality. I pigs SUITABILITY OF DIFFERENT BREEDS FOR DIFFERENT ENVIRONMENTAL 493 CONDITIONS are well-known for their calm temper and especially flavoursome meat. However, they get obese early and have an unappealing external appearance (Tang and Cuong, 1994). Their growth rate is with 200 to 250g/day low. After 8 to 9 months of age, I pigs reach a liveweight of less than 50 kg. The lean meat proportion equals 35 to 39% (Doanh et al., 1985). The reproductive performance of purebred Meo, I and Tap Na is comparable (see Table 4), but lower than that of Mong Cai sows. The litter size of Mong Cai sows is, even under exten- sive conditions, higher than that of (Yorkshire x Mong Cai) crossbreds at station, (Landrace x Mong Cai) crossbreds, and of pure Large White, Landrace and DBI-81. I crossbreds have a higher reproductive performance than pure I but a lower performance than Mong Cai sows. Co pigs have the lowest reproductive performance of all the local breeds.

Table 4: Reproductive performance of different pig genotypes in Vietnam

Production Piglets Piglets Genotype Source system born alive/litter weaned/litter Meo Extensive 7.3 ± 1.5 - Lemke et al., 2002 I Extensive 7.6 ± 2.1 5.6 ± 2.7 Lemke et al., 2000 Co Extensive 5.0 - Hot, 1982 Tap Na Extensive 7.9 - Duc et al., 2004 MC Extensive 12.4 11.5 Lemke et al., 2000 MC Semi-intensive 11.2 ± 2.7 - Lemke et al., 2002 LW - 10.6 - Hot, 1982 LR - 9.3 ± 1.7 7.2 ± 1.61 Thien et al., 1995 DBI-81 - 8.9 ± 1.6 7.3 ± 1.31 Thien et al., 1995 Experimental 9.4 ± 0.6 YR x MC farm 8.6 ± 0.6 Quac et al., 1996 YR x I - 8.3 7.51 NIAH, 2004 LR x I - 8.2 7.21 NIAH, 2004 LR x MC - 10.4 ± 2.4 7.9 ± 1.21 Thien et al., 1995 Pi x I - 13.0 11.0 NIAH, 2004 1 n piglets after 60 days; abbr.: MC = Mong Cai, LW= Large White, YR = Yorkshire, LR = Landrace, Pi = Piétrain Compared to the growth performance of exotic pigs kept at experimental farms, the growth performance of local breeds and its crossbreds with Large White boars under smallholder conditions is poor. Among the local pigs, the Tap Na yielded the highest growth rate, Mong Cai yielded under extensive production conditions a lower growth rate, and the lowest growth rates were found for Meo and (Large White x Meo) genotypes (Table 5). 494 SUITABILITY OF DIFFERENT BREEDS FOR DIFFERENT ENVIRONMENTAL CONDITIONS

Table 5: Live weight gain of different pig genotypes in Vietnam

Genotype Liveweight gain (g/day) Source Meo 186 ± 106 Lemke et al., 2000 Meo (Ban) 65 Lemke et al., forthcoming I 167 Lemke et al., 2000 I 200 – 250 Doanh et al., 1985 Lang Hong 165 ± 29 Lemke et al., 2000 Tap Na 302 Duc et al., 2004 MC 120 Lemke et al., forthcoming MC 166 Valle Zárate et al., 2003 DBI-81 479 Thien et al., 1995 LW 520 Duc et al., 1997 LR 587 Thien et al., 1995 LW x Meo (Ban) 83 Lemke et al., forthcoming LW x Meo 115 Valle Zárate et al., 2003 LW x MC 161 Lemke et al., forthcoming LW x MC 165 Valle Zárate et al., 2003 LW x MC 291 Loc et al., 1996 Abbr.: MC = Mong Cai, LW = Large White, LR = Landrace Summarising the results, local breeds seem to have some unique adaptation characteristics. They have lower performances, which they yield, however, under low-input conditions. In contrast, exotic pigs yield higher performances, but require higher inputs: the higher growth rates in Table 5 were mainly recorded in experimental stations or in breeding centres. How- ever, Mong Cai were shown to have higher reproductive performances than exotic pigs or their crossbreds under improved conditions, and local pig breeds yielded growth rates com- parable to that of (exotic x local) crossbreds. The results in this chapter hint at the production potential of some local pig breeds. However, there have been no systematic investigations on keeping improved breeds under extensive conditions, or keeping local pig breeds under improved conditions. Investigations on the per- formance of crossbreds of different genetic make-up under different production conditions are also missing. Further there is little information on adaptation traits of local pig breeds. Currently, a project by the University of Hohenheim, Institute of Animal Production in the Tropics and Subtropics, in Vietnam is trying to close these gaps by conducting a systematic investigation on different purebred and crossbred genotypes, including exotic breeds (Large White), Vietnamese improved breeds (Mong Cai) and local breeds (Meo/Ban) under differ- ent production conditions in a cross-classified design. Further studies deal with the impact SUITABILITY OF DIFFERENT BREEDS FOR DIFFERENT ENVIRONMENTAL 495 CONDITIONS that (governmental) subsidies have on farmers’ keeping local or exotic pig breeds in Viet- nam, and explore market niches for pork produced from local breeds. Over the last decades, exotic pigs have steadily spread and replaced local pigs in the Viet- namese pig population (see also Figure 5). As part of ongoing socio-economic developments, these exotic pigs have become increasingly available and accessible for farmers and have enabled them to produce pork with increasing efficiency. However, the question remains whether pig-keeping resource-poor smallholders in remote and mountainous regions can be included, or if they can set up niche production with local pig breeds. 496 IMPACT OF IMPORTS ON BIODIVERSITY

7 IMPACT OF IMPORTS ON BIODIVERSITY The third edition of the World Watch List for Domestic Animal Diversity published by the FAO and UNEP reports that every week the world loses two breeds of domestic animals, while 1,350 breeds face extinction in the near future. The greatest threat to domestic animal diversity is the export of animals from developed to developing countries, leading to cross- breeding and even replacement of local breeds. In developing countries, breeds from the in- dustrialized world are considered more productive. The problem, however, is that those ani- mals realize performances according to their genetic potential only under management condi- tions and at an input level that most local farmers cannot supply, thus putting them under a considerable economic risk. For a long time, the animal production of Vietnam was based on local genotypes. Due to their slow growth, low feeding efficiency and early deposit of fat, local genotypes have been progressively “improved” through crossbreeding or been replaced with imported high- yielding breeds (“lean meat programs”). As shown above, Vietnam has a long history of im- porting exotic pig breeds; under French and American rule and later on as the Socialist Re- public of Vietnam. At certain earlier times, the government had supported higher-yielding local breeds, especially the Mong Cai (“Mong Cai-isation”). Importing exotic breeds and promoting a single higher-yielding local breed have led to a severe decrease in the number of indigenous breeds. Meanwhile, exotic pigs and their crossbreds dominate the Vietnamese pig production, and some local breeds have already disappeared or are severely endangered by extinction (NIAH, 1997). As an example, in North Vietnam the proportion of local genotypes in the total sow population decreased from 72% in 1994 (Thien et al., 1996) to 45% in 1997 (NIAH, 1997). In 2002, the Vietnamese pig population was a mere 26% of local pigs. Among 14 indigenous Vietnamese pig breeds, five breeds were in vulnerable state (36%), two in critical state (14%), and three were facing extinction (21%) (Country Report of Viet- nam, 2003). To prevent the extinction of local breeds, Vietnam founded the National Program on Conser- vation of the Vietnamese Animal Genetic Resources. Under the direction of NIAH, special sub-programs were founded for the protection and conservation of certain domestic animal breeds. In contrast to the majority of conservation programs in other countries, most of the conservation programs implemented by NIAH are conducted on farms with farmer participa- tion (Lemke et al., 2000). There is a growing recognition that preserving local breeds is not only important to ensure the livelihoods of resource-poor farmers depending on those animal breeds; but their conservation is regarded as a national insurance policy, as locally adapted animal genetic resources could become future assets in livestock breeding programs. According to the FAO Global Databank for Farm Animal Genetic Resources, most Vietnam- ese local pig breeds are not at risk, and population trends are stable (Mong Cai) or decreasing (e.g. Co, Lang Hong, I, Meo). Only the Bo Xu is considered extinct. However, according to later Vietnamese sources, those data do not reflect real conditions: I pigs were in the 1970s and 1980s widely kept in the North Vietnamese delta provinces (Ly, 1999) and used to create improved crossbred genotypes (see above). As Mong Cai pigs have progressively replaced the I pig as sow line (Ly, 1999), the I population was at the edge of extinction at the end of the 1970s (Thuy, 1996). It is nowadays in a critical condition, with a decreasing population trend (Country Report of Vietnam, 2003). IMPACT OF IMPORTS ON BIODIVERSITY 497

The Mong Cai was for a long time the dominating breed in North and Central Vietnam, and its population increased quickly between the 1960s and 1980s. Initially, the government sup- ported its spread and use nationwide. Since 1975, state farms exported Mong Cai to other provinces, to replace lower yielding local pig breeds (Duyet and Duong, 1996). In the course of its ubiquitous distribution, the Mong Cai has mixed with Lang Hong pigs and other local pig breeds (Doanh et al., 1985; Ly, 1999). Although Mong Cai were raised at state farms and widely promoted, the population declined in the early 1990s. After an increase in population size up to 1995, the population is now stable, but the population trend is decreasing, and the degree of crossbreeding in the population increasing (Country Report of Vietnam, 2003). The Mong Cai remains one of the major local sow lines in North Vietnam (Ly, 1999). In North Western mountainous provinces, the Meo (and related breeds/types) are commonly kept. However, exogenous and indigenous influences may lead to a replacement of local Meo/Ban pigs (Lemke et al., 2002). Among others, the construction of a huge hydropower plant in Son La province and related effects concerning transportation, infrastructure, con- nectedness to markets, and income of inhabitants, will probably influence the keeping of local Meo/Ban pigs in and around Son La. Meo and Muong Khuong are currently not at risk, while Ban and H’mong are in a vulnerable state. For all four breeds, populations are declin- ing. The Lang Hong is currently in a critical state, with a decreasing population trend. Ba Xuyen and Thuoc Nhieu are in a vulnerable state, also with declining population. The Phu Khanh faces extinction; and the pure population is decreasing due to mixing with other breeds. In Central Vietnam, the Soc is not at risk, while the Co is already facing extinction. For both, populations are declining. In the last survey, no pure Co boars were found anymore. The Son Vi faces extinction, the Mini pig of Quang Tri is in a vulnerable state with declining popula- tion (Country Report of Vietnam, 2003). Genetic diversity supports livestock’s ability to adapt to many unfavourable environmental factors like diseases and parasites, variations in the availability and quality of feed and water, and extreme climate conditions. Animals in developed countries increasingly belong to a small number of high-performance breeds and hybrids, which have been developed over the last two centuries, strongly influenced by controlled scientifically funded breeding programs. Those animals have been selected for high yields, not for adaptation, and require standard- ised conditions and high inputs for exploitation of their potential. Vietnamese local breeds are specific for particular regions, representing a large natural gene pool. Compared to exotic breeds, they show a high genetic diversity, although the Mong Cai breeds do so to a lesser extent. Vietnamese local breeds differ genetically according to their geographic location (Thuy, 2004). However, as the example of Meo/Ban highlights (see above), the definition of local pig breeds in Vietnam is not fully standardised. Indigenous populations are a source of adaptability for specific environmental challenges such as disease and extreme climatic conditions and a reservoir of worldwide genetic diversity for possible future changes in production systems (Olivier et al., 2002). Thuy (2004) showed that Viet- namese indigenous breeds were genetically distant to European pig breeds, had a higher number of alleles per gene locus, wider ranges of allelic sizes, and were genetically more heterogeneous than European breeds. The large genetic distance between the Vietnamese and European breeds could be exploited in crossbreeding, benefiting from heterosis and combina- tion effects for performance traits. Vietnamese pigs harbour also a source of new alleles, 498 IMPACT OF IMPORTS ON BIODIVERSITY which might be significant for future genetic improvement and of unpredictable economic value. To which degree that prospective source will be maintained, depends on national will- ingness to pay for conservation programs. This depends on expectations of future benefits. Those expectations will not be based on fairy tales on the overall goodness of local breeds, but only on scientific proofs of special value of specific traits, controlled by specific alleles, and market-backed values of products. CONCLUSIONS 499

8 CONCLUSIONS Vietnam owns a considerable variety of local pig breeds. The introduction of pigs and breeds from neighbouring countries (Laos, Cambodia, China) started probably centuries ago, as part of human migration (e.g. Thai and H’mong migrating from China), occupation (China), and trade. The influx of breeds was an important component in the development of Vietnamese local breeds. However, information is lacking on those early phases. The earliest confirmed information on pig breed introduction goes back to the 1920s. Gene flow in the recent past and present has probably been a net inflow of pigs. Exports (e.g. Vietnamese potbellied pigs to western countries as pets and for scientific use) were negligi- ble. Before 1955 (end of French colonisation) and after 1986 (economic liberalisation), pig imports were directed by commercial interests as the main driving force of gene flow. From 1955 until 1986 the major driving force was the policies of the socialist government, and after 1990 additionally foreign developmental projects, both with the declared aim to benefit the poor farmers, but not always fulfilling their claim. The inflow of pig breeds to Vietnam consisted of higher-yielding breeds from Europe and America, which were introduced due to their higher performances (in the countries of origin) to improve or replace the low yielding local breeds. Commercial imports consisted of exotic pigs. Current development and poverty alleviation projects at village level usually promote exotics, and only occasionally improved Vietnamese breeds (e.g. promotion of Mong Cai by Vétérinaires sans frontières, Phu Tho). Information on pig gene flow to and within Vietnam is limited, due to the restricted informa- tion policy of both international breeding companies and Vietnamese official sources, but also due to the decentralised nature of pig breed import and distribution. The introduction of exotic pigs was supported by the decentralised nature of the Vietnamese breeding system. Centralised coordination of breeding measures is not well developed, and centralised measures fulfilled their aims only partly. However, the impact of the state-run breeding stations has been considerable; and the advanced use of AI has strongly supported the introduction of exotic genetics to the smallholder producer level. The influx of exotic breeds has positively influenced output and efficiency of pork produc- tion in Vietnam, while the local pig populations have been reduced. Today, pigs of various crossbreeding degrees are widely distributed. Most indigenous breeds show declining popu- lation trends, and the majority of local breeds are in a vulnerable or critical condition or even facing extinction. Conservation measures of Vietnamese institutions follow suitable ap- proaches (in-situ conservation on-farm). However, due to shortcomings in set-up and imple- mentation, they may not successfully preserve local pig breeds. National decisions and the willingness to pay for conservation programs depend on expectations for future benefits, which need to be based on scientific proofs of the value of specific traits, and market-backed valuations of products. Research results indicate a considerable production potential of local pig breeds especially under low-input conditions, favourable adaptation traits, and genetic peculiarities, differenti- ating them from the European breeds. Local pig breeds are a significant component of the Vietnamese and worldwide biodiversity, are important for resource-poor farmers in Vietnam who depend on local breeds to ensure their livelihoods, and for future breeding measures 500 CONCLUSIONS utilising e.g. favourable adaptation traits. On the other hand, exotic pigs have become in- creasingly available and accessible for farmers in Vietnam and have enabled them to produce pork with increasing efficiency. Whether pig-keeping resource-poor smallholders in remote and mountainous regions can be integrated in this process, or if they can set up niche produc- tion with local pig breeds, remains to be clarified by further investigations. Further investiga- tions are required to define local pig breeds, further characterise their genetic specificities, and to comparatively evaluate their performances under standardised conditions. REFERENCES 501

9 REFERENCES ACIAR 2004. Research that works for developing countries and Australia. Project: Breeding and feeding pigs in Australia and Vietnam. Anh, N. Q. and H. S. Dung. 1994. Lon Meo-Nghe An. Meo pigs-Nghe An. p. 105. In: Ket qua nghien cuu bao ton quy gen vat nuoi 1990-1993 (Results of the research on con- servation of domestic animal genetic resources 1990-1993). NIAH Ministry of Sci- ence, Technology and Environment; Ministry of Agriculture and Industry, Hanoi, Vietnam. (in Vietnamese). ASPS. 2002. Livestock policy briefs for Vietnam. MARD - DANIDA: Agriculture Sector Programme Support ASPS; Agrifood Consulting International ACI, Hanoi, Vietnam. Astroem H. 2000. The Mongcai pig: digestibility studies of fibrous feeds, and a survey of smallholder production systems in two agro-ecological zones of Vietnam. Swedish University of Agricultural Sciences, Sweden. Bosma, R. H., R. L. Roothaert, P. Asis, J. Saguinhon, L. H. Binh, and V. H. Yen. 2003. Eco- nomic and social benefits of new forage technologies in Mindanao, Philippines and Tuyen Quang, Vietnam. Centro Internacional de Agricultural Tropical (CIAT), Los Baños, Philippines. Country Report of Vietnam for FAO’s State of the World’s Animal Genetic Resources. 2003. Doanh, P. H. 1985. Ket qua khao sat, giam dinh va dieu tra giong lon o mien Nam (Research results on pig breeds in the South). p. 7. In: NIAH (ed.). Studies on technique and science 1969-1979. Agricultural Publishing House. Hanoi, Vietnam (in Vietnamese). Doanh, P. H. 1994. Bao ton von gen quy cua giong lon Mong Cai (Conservation of the valu- able gene resource of the Mong Cai breed). p. 30. In: NIAH (ed.). Ket qua nghien cuu bao ton nguon gene vat nuoi o Viet Nam (Results of livestock genetic conservation in Vietnam). Agriculture Publishing House, Hanoi (in Vietnamese). Doanh, P. H. et al. 1985. Mot so dac diem ve tinh nang san xuat cua giong lon noi (Produc- tion performance of local pig breeds). p. 10. In: Tuyen tap cong trinh nghien cuu chan nuoi 1969-1984 (Research in animal husbandry 1969-1984). NIAH. Hanoi, Vietnam. (in Vietnamese). Doanh, P. H. and D. H. Luan. 1985. Ket qua nghien cuu dac diem sinh hoc va tinh nang san xuat cua mot so giong lon ngoai (Research results on the production performance of some exotic pig breeds). p. 19. In: NIAH (ed.). Tuyen tap cong trinh nghien cuu chan nuoi 1969-1984 (Research in animal husbandry 1969-1984). Hanoi, Vietnam. (in Vietnamese). Doanh, P. H. and T. T. Thong. 1985. Ket qua nghien cuu lai tao giong lon moi DBI-81 va BSI-81 (Research results on the new crossbred pig breeds DBI-81 and BSI-81). p. 28. In: Tuyen tap cong trinh nghien cuu chan nuoi 1969-1984 (Research in animal hus- bandry 1969-1984). Hanoi, Vietnam. (in Vietnamese). Dong, N. V. and P. S. Tiep. 2002. Ket qua phat trien giong lon Mong Cai o Tuyen Quang. (Developping the Mong Cai pig in Tuyen Quang province). Khoa hoc ky thuat chan nuoi (Information on science and technologies in animal production) 5: 14-19 (in Vietnamese). 502 REFERENCES

Duc, N. V., B. P. Kinghorn, and H. U. Graser 1997. Studies on some Production and Carcass Traits of Mong Cai and Their Crosses in North Vietnam. p. 185. In: Proceedings of the Association for the Advancement of Animal Breeding and Genetics. Proceedings of the 12th Conference. RSL Complex and Western Plains Zoo Dubbo, NSW, Aus- tralia; April 6 to 10, 1997. Australia. Duc, N. V., G. H. Tuyen, and D. C. Tuan. 2004. Mot so dac diem co ban cua giong lon Tap Na (Basic characteristics of the local Tap Na pig breed in Thong Nong, Cao Bang). Download from http://www.vcn.vnn.vn/qg/quygen2004/qg_20_11_2004_2.htm at December 20, 2004. (in Vietnamese). Dung, T. V. 2002. The current status of veterinary service in Vietnam and development ten- dency aims to commercial animal husbandry in the first decade of 21st century. Workshop organized by NIAH and CIRAD, Hanoi, October 24-25, 2002. Hanoi, Vietnam. Duy, P. V., N. X. Hoan, and V. T. Binh. 2001. Ket qua buoc dau xay dung mo hinh nuoi lon nai ngoai giong Yorkshire trong dieu kien chan nuoi nong ho o Hop Tien-Nam Sach- Hai Duong (Creating a model to raise Yorkshire sows under smallholder conditions in Hop Tien, Nam Sach, Hai Duong province). In: Bao Cao khoa hoc chan nuoi thu y 1999-2000, phan chan nuoi gia suc (Scientific report on veterinary medicine and animal husbandry 1999-2000, animal husbandry part). pp. 67-82. Hanoi, Vietnam. (in Vietnamese). Duyet, H. N. and N. K. Duong. 1996. Pig breeding in central Vietnam and its improvement. In: Exploring approaches to research in the animal sciences in Vietnam: A workshop held in the city of Hue, Vietnam 31 July – 3 August, 1995. Canberra, Australia. FAO. 1999. Livestock in food security in Vietnam. RAP Publication. FAO. Bangkok, Thai- land. FAO. 2004. DAD-IS. Download from http://www.fao.org/dad-is at November 29, 2004. General Statistical Office Vietnam. 2000. Statistical data of Vietnam. Agriculture, forestry and fishery, 1975 to 2000. Statistical Publishing House. Hanoi, Vietnam. Hai, L. T. 1994. Lon Thuoc Nhieu, mot nguon gene quy o dong bang song cuu long (Thuoc Nhieu pig, a valuable gene resource in the Mekong Delta). p. 42. In: Ket qua nghien cuu bao ton nguon gene vat nuoi o Viet nam (Results of livestock genetic conserva- tion in Vietnam). Agriculture Publishing House, Hanoi, Vietnam. (in Vietnamese). Hai, L. T. 1996. Social features and the development of swine production in Vietnam. ACIAR Proc. 68: 186 – 189. Hai, L. T. and N. H. Nguyen. 1997. Outlines of pig production in Vietnam. Pig News and Information. 18: 91 – 94. Hoa, P. T. T. forthcoming. Phenotypic and Biometric Characterisation of Autochthonous Vietnamese Ban Pigs. MSc thesis. University of Hohenheim, Germany. Hot, V. T. 1982. Mot so dac diem ngoai hinh va sinh truong cua cac giong lon o Tay nguyen (Features of outer appearance and performance of pig breeds in the Central High- lands). Khoa hoc va ky thuat nong nghiep (Agricultural technique and science maga- zine) 13: 124 – 126. (in Vietnamese). Huy, V. S., L. V. Ly, and H. K. Giao. 1996. Mot so ket qua nghien cuu buoc dau ve sinh truong sinh san cua dan lon ngoai nuoi tai xi nghiep lon giong Dan Quyen Thanh Hoa (Performance of exotic pig breeds raised in a breeding enterprise in Thanh Hoa prov- REFERENCES 503

ince). p. 24. In: Results of technical and scientific studies on animal husbandry 1994- 1995. Agricultural Publishing House, Hanoi, Vietnam. (in Vietnamese). Huyen, L. T. T. 2004. Analysis of animal production systems in H’mong villages of North- ern Vietnam. MSc thesis. University of Hohenheim, Germany. IFAD. From http://www.ifad.org/evaluation/public_html/eksyst/doc/profile/pi/vn.htm, April 1, 2005. Khanh, N. and D. D. Hien. 1963. Buoc dau khao sat ve giong lon I (First research on the I). Khoa Hoc va Ky Thuat Nong Nhgiep (Science and Technique in Agriculture) 11:675- 683. (in Vietnamese). Konrad-Adenauer-Stiftung. Vietnam Office. From http://www.kas.de/proj/home/home/72/2/, download at April 1, 2005. Lemke, U., L. T. Thuy, A. Valle Zárate, and B. Kaufmann. 2000. Characterisation of a model for conservation of autochthonous pig breeds on smallholder farms in North Vietnam. Project report for GTZ/ Sectoral Project: Tropical Ecology Support Program TOEB. Stuttgart, Germany. Lemke, U., L. T. Thuy, A. Valle Zárate, and B. Kaufmann. 2002. Characterisation of small- holder pig production systems in mountainous areas of North Vietnam. Deutscher Tropentag 2002. October 9 to 11, 2002, Kassel-Witzenhausen, Germany. Lemke, U., L. T. Thuy, B. Kaufmann and A. Valle Zárate. Forthcoming. Evaluation of smallholder pig production systems in North Vietnam considering input, manage- ment, output and comparing economic and biological efficiency. Lich, L. B. 1996. Livestock production of Vietnam in the passing year and development ori- entations toward 2000. MARD & VUSTA – National Seminar on Animal Science and Development to the year 2000. 26 – 28 November, 1996. Animal Husbandry as- sociation of Vietnam, Hanoi, Vietnam. Lich, L. B. and D. K. Tuyen. 2001. Livestock production development in the Northern prov- inces of Vietnam. Khoa hoc ky thuat chan nuoi- Information of science and technolo- gies on animal production. Hanoi, Vietnam. (in Vietnamese). Loc, N. T., B. Ogle, and T. R. Preston. 1996. Pig production in Central Vietnam: results of a participatory rural appraisal survey and on-farm feeding trials with protein supple- mentation of traditional diets. Integrated farming in Human development. Proceed- ings of a workshop. March 25 – 29, 1996. Tune Landboskole, Denmark. Luong, T. V. and H. Gian. 1999. Kha nang sinh san cua lon Nai F1 (Yorkshire x Mong Cai) nuoi o ho gia dinh tai huyen yen dinh, Thanh Hoa (High reproduction of Yorkshire x Mong Cai sow raising in Yen Dinh, Thanh Hoa province). Tap chi chan nuoi (Animal husbandry magazine) 4 (25) 14 – 17 (in Vietnamese). Ly, L. V. 1993. Note on the local animal genetic resources and their conservation in Viet- nam. Animal genetic resources information. UNEP, Rome, Italy. Ly, L. V. 1996. A review of animal science research in Vietnam. In: Exploring approaches to research in the animal sciences in Vietnam: A workshop held in Hue, Vietnam 31 July – 3 August, 1995, Canberra, Australia. Ly, L. V. 1999. Chuyen khao: Bao ton nguon gen vat nuoi o Viet nam (Animal genetic con- servation in Vietnam). Agriculture Publishing House, Hanoi, Vietnam. (in Vietnam- ese). 504 REFERENCES

Ly, L. V. 2000. The development of a sustainable animal production system based on the advantage of tropical agriculture. Workshop-seminar ''Making better use of local feed resources'' January, 2000. SAREC-UAF, Hanoi, Vietnam. Ly, L. V. and T. T. B. Duyet. 2000. Mong Cai pig breed – a good genetic resource in Viet- nam. Hanoi, Vietnam. Not published. Minh, L. H. 2000. Ket qua 6 nam (1992-1998) thuc hien Mong Cai hoa dan lon nai nen o Tuyen Quang (Six years Mong Cai-ization in Tuyen Quang province). Tap chi chan nuoi (Animal husbandry magazine) 2:16-18. (in Vietnamese). Molenat, M. and T. T. Thong. 1991. La production porcine au Viet Nam et son amélioration. WAR 68:26-36. (in French). NIAH. 1985. Ket qua nghien cuu chan nuoi lon (Research on pig husbandry). p. 7. In: Re- search on technique and science from 1969-1979. Agricultural Publishing House, Hanoi, Vietnam. (in Vietnamese). NIAH. 1997. The national policy on conservation and use of indigenous animal genetic re- sources in Vietnam. GCP/RAS/144/JPN project. Hanoi, Vietnam. NIAH. 2003. Livestock research and development towards 2010. A strategic plan for the National Institute of Animal Husbandry (NIAH), Vietnam. Information on science and technologies in animal production. National Institute of Animal Husbandry. Ha- noi, Vietnam. NIAH. 2004. Ket qua buoc dau nghien cuu cac cong thuc lai giau lon nai i va lon duc ngoai (Results of crossbreeding between I sows and exotic boars in Thanh Hoa province). Download from http://www.vcn./qg/quygen2004/qg_20_11_2004_2.htm, December 20, 2004. (in Vietnamese). Nhien, T. 2004. UK holds pig breeding conference in Vietnam. Vietnam Economy. http://www.vneconomy.com.vn/eng/article_to_print.php?id=041208104058, April 1, 2005. Olivier, J. J., S. Moyo, H. H. Montaldo, W. Thorpe, A. Valle Zárate, and K. R. Trivedi. 2002. Integrating genetic improvement into livestock development in medium-to low-input production systems. 7th World Congress on Genetics Applied to livestock produc- tion. August 19-23, 2002. Montpellier, France. Peters, D. 1998. Improving small-scale pig production in northern Vietnam. WAR 91: 2– 12. Pig International. 1998. Foreign investors speed Vietnam's transition. Volume 28(1): 25 – 27. Quac, N. K., T. V. Phung, and G. Husssain Jaffar. 1996. Study on the physiological charac- teristics and reproduction of crossbred-sows F1 (Yorkshire x Mong Cai). Agricultura Tropica et Subtropica. Universitas Agriculturae Praga 29: 59 – 64. Rodríguez, L. J., T. R. Preston, and F. Dolberg. 1996. Participatory Rural Development: ''Experiences in Binh Dien and Xuan Loc villages in Central Vietnam''. Livestock Re- search for Rural Development 8: 1 – 40. Singh, D. N., C. P. McPhee, and J. S. Kopinski. 1996. Helping to improve pig production in Vietnam. ACIAR Proc. 68: 176 - 181. Canberra, Australia. Tang, D. X. and N. N. Cuong. 1994. Ket qua buoc dau giu quy gen lon I Thanh Hoa (Results of genetic conservation of I pigs in Thanh Hoa). In: NIAH (ed.) Ket qua nghien cuu bao ton nguon gene vat nuoi o Viet nam (Results of the livestock genetic conserva- tion in Vietnam). Agriculture Publishing House, Hanoi, Vietnam. (in Vietnamese). REFERENCES 505

Thien, N. 2002. Ket qua nghien cua va phat trien lon lai co nang suat va chat luong cao o Viet nam (Research results on developing crossbred pigs with high quality and high yield in Vietnam). p. 82. In: NIAH (ed). Vien Chan nuoi 50 nam xay dung va phat trien (NIAH - 50 years of development). Agriculture Publishing House, Hanoi, Viet- nam. (in Vietnamese). Thien, N., P. H. Doanh, and P. T. Van. 1992. Kha nang sinh san cua cac giong lon Landrace, Dai Bach, DBI-81 va cac cap lon lai huong nac (Reproductive performance of Land- race, Large White, DBI-81 and other lean meat hybrid pigs). p. 17. In: Results of technical and scientific studies on animal husbandry from 1985-1990. Agricultural Publishing House, Hanoi, Vietnam (in Vietnamese). Thien, N., P. T. Van, N. K. Quac, P. H. Doanh, and P. N. Le. 1995. Ket qua nghien cuu cac cong thuc lai giua lon ngoai va lon Vietnam (Studies on crossbreeding between exotic and Vietnamese pig breeds). Selection of scientific works on animal production 1969- 1995. Hanoi, Vietnam. (in Vietnamese). Thien, N., P. T. Van, P. N. Le, P. H. Doanh, N. Nghi, N. K. Quac, and V. T. Hot. 1996. Im- provement of productivity and meat quality of pigs in the Red River Delta region by crossbreeding. ACIAR, Canberra, Australia. Thong, T. T. 1996. Animal husbandry advances which contributed to major livestock devel- opment in Vietnam. In: Exploring approaches to research in the animal sciences in Vietnam: A workshop held in Hue, Vietnam 31 July – 3 August, 1995, Canberra, Australia. Thong, T. T., L. T. Hai, and V. A. Quac. 1996. Research wok on pig production in the Me- kong Delta. ACIAR Proc. 68: 56. Australia. Thu, A. 2004. Lon Muong Khuong se xuong nui. Lao Dong newspaper vol. 312 (in Viet- namese). Thuong, N. V. 1985. Vien chan nuoi, 15 nam nghien cuu khoa hoc va san xuat 1969-1984 (NIAH – 15 years of research and production 1969-1984). NIAH, Hanoi, Vietnam (in Vietnamese). Thuy, L. T. 1999. Lon Meo Nghe An (Meo pigs in Nghe An). NIAH, Hanoi, Vietnam. Not published (in Vietnamese). Thuy, N. T. 1999: Desk-study on animal health field services in Vietnam (SVSV ALA/96/20). Ministry of Agriculture and Rural Development, European Commis- sion, Vietnam. Thuy, N. T. D. 2004. Genetic diversity and distances of Vietnamese and European pig breeds analysed with microsatellite loci. Dissertation. Shaker Verlag. Aachen, Germany. Tjällden, U. 1999. Indigenous breeds of domestic animals in northern Vietnam. Opinions of, and attitudes towards, local breeds among farmers. Minor Field Studies International Office, Swedish University of Agricultural Sciences, Uppsala, Sweden. To, L. V. and L. D. Duc. 1967. Giong lon Meo nuoi o Dong Bang- Nghe An (Raising Meo pigs in Dong Bang, Nghe An province). Tap chi khoa hoc ky thuat nong nghiep nam 1967 (Journal of Science and Technology in agriculture 1967) 17: 154 – 158 (in Vietnamese). TOPIGS. Download from http://www.topigs.ca/pbcd.htm at April 1, 2005. Truc, D. X., N. Thien, N. H. Cuong, T. V. Linh, P. T. Van, P. S. Tiep, N. H. Tao, and P. S. Lang. 2003. So tay ky thuat chan nuoi lon trang trai (Handbook of the technique in 506 REFERENCES

pig husbandry on farm). Agricultural Publishing House, Hanoi, Vietnam (in Viet- namese). Tung, D. X. 1999. Roles of livestock on smallholder farming systems in northern provinces of Vietnam. Agricultural University of Norway; Centre for International Environment and Development Studies, Norway. Valle Zárate, A., B. Kaufmann, U. Lemke, L. T. Thuy, L. V. Ly, and N. D. Vang. 2003. Effi- ciency of smallholder animal husbandry depending on intensity of management and genetic potential of livestock in mountainous regions of Northern Vietnam. Interim report to DFG, Germany. Universität Hohenheim, Stuttgart, Germany. Vang, N. D. 2002. Livestock Development of Vietnam towards 2010: prospect and chal- lenges. p. 71. In: NIAH (ed.) Vien Chan nuoi 50 nam xay dung va phat trine (Na- tional Institute of Animal Husbandry - 50 years of development). Agricultural Pub- lishing House, Hanoi, Vietnam (in Vietnamese). Xuan, V. T., L. Th. Hai, and Ch. B. Loc. 1995. Research priorities for improving animal pro- duction by agro-ecological zone in Vietnam. p. 197. In: IRRI (ed.) Global Agenda for Livestock Research. Proceedings of the consultation for the South-East Asia region, 10 - 13 May 1995, IRRI, Phillipines. CONTACT ADDRESSES 507

10 CONTACT ADDRESSES Le Thi Thanh Huyen Regina Roessler Institute of Animal Production in the Tropics Institute of Animal Production in the Tropics and Subtropics and Subtropics University of Hohenheim University of Hohenheim 70593 Stuttgart 70593 Stuttgart Germany Germany Phone: +49 (0) 711 459 3006 Phone: +49 (0) 711 459 3006 Fax: +49 (0) 711 459 3290 Fax: +49 (0) 711 459 3290 Email: [email protected] Email: [email protected] Ute Lemke Prof. Dr. Anne Valle Zárate Institute of Animal Production in the Tropics Institute of Animal Production in the Tropics and Subtropics and Subtropics University of Hohenheim University of Hohenheim 70593 Stuttgart 70593 Stuttgart Germany Germany Phone: +49 (0) 711 459 3294 Phone: +49 (0) 711 459 3170 Fax: +49 (0) 711 459 3290 Fax: +49 (0) 711 459 3290 Email: [email protected] Email: [email protected] Dr. Le Thi Thuy Dr. Nguyen Van Dong Animal Genetic Molecular Laboratory Pig research centre National Institute of Animal Husbandry National Institute of Animal Husbandry Chem, Tu Liem Chem, Tu Liem Hanoi, Vietnam Hanoi, Vietnam Phone: +84 (4) 838 9165 Phone: +84 (4) 938 9774 Fax: +84 (4) 8389775 Mobile: +84 (9) 13001340 Email: [email protected]

508