The concept of a breed, in which all the members have a pedigree tracing their ancestry, was developed primarily in western Europe during the eighteenth century. Today, in the developed world, breeds are recognized as distinct intraspecific groups, the members of which share particular characteristics, which distinguish them from other such groups, and formal organizations usually exist for each breed or breed group. The term breed as a formal designation has little meaning outside areas of Western influence, where pedigree recording is often non-existent. Nonetheless, even under these circumstances, there exist strains or "types", which owe their continuing distinct identity to a combination of traditional "breeding objectives" and geographical and/or cultural separation by the communities that own them. This paper discusses the concept of livestock breeds in the context of community-based management of farm animal genetic resources in developing countries. It is pointed out that elaborate traditional systems of population identification by local communities exist, and that these "uniquely identified and named" populations, the equivalent of "breeds" in Western agriculture, are important basic units of diversity assessment and conservation. In many instances, the local environment, culture and values of communities are reflected in the key traits and/or functions of the livestock "breeds". These may range from such complex traits as adaptation to local environmental stresses, to functions such as traction or products such as milk or wool and simple traits such as presence and size of horns or humps. Behind these traits are the breeding practices and other indigenous knowledge systems that have been applied for the maintenance of these populations, and the genetic diversity that they represent. The paper concludes that the concept of breed (or equivalent concepts within species) serves an important purpose in that it links products/functions to a group of animals that share a common genetic background. To the extent that conservation of agricultural diversity needs to be linked to utilization, and given that the concept of breeds as aggregate or "package" of traits is a manifestation of the environment and community values and goals, breeds represent the single most important unit of analysis in the context of conservation and use of livestock diversity. Moreover, they must be considered as such in developing strategies for the sustainable management of livestock diversity at the community level.
There are two broad approaches through which farm animal genetic resources (AnGR) can be conserved: ex situ and in situ. The ex situ approach to conservation includes methods such as cryopreservation and live-animal conservation in designated localities, e.g. government farms. In situ conservation, also referred to as "on-farm conservation", can be defined as "the continuous husbandry of a diverse set of populations by farmers in the agro-ecosystems where an animal population/breed/strain has evolved". On-farm conservation encompasses entire agro-ecosystems, including immediately useful species (of crops, forages, agroforestry species, other animal species) that form part of the system. There is a wide range of objectives that may underpin an on-farm conservation programme. These may include the following:
to conserve the processes of evolution and adaptation of animal populations to their environments;
to conserve diversity at all levels - ecosystem, species and within species (breeds and genes);
to integrate farmers (mixed farmers, pastoralists) into a national AnGR system;
to conserve ecosystem services that are critical to the functioning of the Earths life-support system (maintaining soil-forming processes, reducing chemical pollution, restricting the spread of animal and plant diseases, etc.);
to improve the livelihood of resource-poor farmers through economic and social development - e.g. combining on-farm conservation with development of local infrastructure, or increasing access by farmers to useful locally relevant animal and plant (forage) germplasm;
to develop systems to make conserved material (e.g. semen for local use) or conditions easily accessible to farmers.
There are several advantages of in situ conservation of AnGR. One advantage is that it conserves both the genetic material and the processes that give rise to the diversity. Blackburn et al. (1996) have provided examples that illustrate the fact that those traditional breeds with a history of interaction with wildlife evolve an interlocking pattern of vegetation exploitation so that the pasture can support a maximum of biomass. Thus, adapted indigenous breeds can be co-conserved with wild species, maximizing system output sustainably. Long-term sustainability of breeding efforts may depend on the continued availability of the genetic variation that can be maintained and further developed by the herders themselves using their own management practices. Moreover, because the technology for cryopreservation of AnGR is only well-developed for a handful of livestock species, conservation of most livestock species will, of necessity, continue to depend on live animals. In almost all cases, interventions supporting continued evolution (in response to changes in the production system) is cheaper and more effective for AnGR conservation on-farm. In the case of ex situ approaches, there would need to be regeneration and exposure to the environment followed by sampling and preservation of a subset. Referring to the situation in plant genetic resources conservation, Brush (1991) has pointed out that "long-term sustainability of breeding efforts may depend on continued availability of larger amounts of germplasm than can be effectively stored off-site". Rephrased in the context of AnGR, conserved animal material in ex situ systems is more likely to be utilized in emergency restoration but is much less likely to find use in long-term animal improvement programmes.
Unfortunately, in situ conservation also has some drawbacks. The first one is that the same factors that allow for dynamic, holistic, agro-ecosystem conservation may serve to threaten the security of breeds/strains. For example, genetic erosion can still occur as a result of unforeseen circumstances such as war and natural disasters. Moreover, social and economic change may either foster or hinder on-farm AnGR conservation over time. Indeed, one of the challenges of in situ conservation research is to evaluate how economic development is affecting farmer maintenance of diversity so as to take account of this process in the implementation of conservation programmes. Because each of the two broad conservation approaches has its merits and demerits, there is need for an "integrated conservation approach" that combines a range of available ex situ and in situ options.
Community-based management and in situ conservation of AnGR
Why community-based conservation? The fact that the role of community-based conservation has received increasing attention derives from the realization that most creative and productive activities of individuals or groups in society take place in communities. Communities and citizens groups provide the easiest means for people to take socially valuable action as well as to express their concerns. When they are properly empowered and informed, communities can contribute to decisions that affect them and play an indispensable part in creating a sustainable society (IUCN/UNEP/WWF, 1991). As local communities have a vested interest in all the natural resources (including AnGR) on which their livelihoods depend, and have the most to lose in the event of loss of these resources, they are best placed to conserve them. Moreover, they have a better understanding than any other group of what it takes to manage their traditional resources sustainably.
What is the meaning of community? During the international negotiations that produced the Convention on Biological Diversity (CBD), such phrases as "indigenous people", "indigenous knowledge" and "traditional knowledge" were often used in reference to local or traditional "communities", but the definition of some of these terms remains problematic. The term "indigenous" is particularly problematic because it has different meanings in different parts of the world. "Local communities" is really a broad term that has no connotation of "nativity", but refers generally to people who, at a particular point in time, have common interests and live in a defined geographical area, rural or urban, within a broader society. Thus, when the concept of "indigenous people" is broadened to include local communities, "embodying traditional lifestyles", it forms the basis for a broad and strong alliance of those peoples who still "live close to the Earth" (IUCN/UNEP/WWF, 1991). These peoples embody the wisdom of generations that have practised the sustainable lifestyles that environmentalists and conservationists promulgate. Thus, "community-based" management of AnGR refers to a system of AnGR and ecosystem management in which the AnGR keepers are responsible for the decisions on definition, priority setting and the implementation of all aspects of conservation and sustainable use of the AnGR. The community does not have to be "indigenous" in its strictest definition, neither should the lifestyles remain strictly traditional. Indeed, the dynamic processes that characterize the evolution of agricultural systems necessitate that the communities should themselves also be dynamic in their aspirations and strategies for managing their livelihoods. Community-based development approaches recognize, and respond appropriately to, the dynamism of the systems involved.
As defined above, in situ conservation is the management of viable populations (by farmers) in the agro-ecosystems where they have developed their distinctive properties. Thus, "in situ conservation" and "community-based management" of AnGR are conceptually similar. However, there are subtle, but significant differences. "Conservation of AnGR" has been defined (FAO, 1992) as the sum of all actions involved in the management of AnGR, such that these resources are best used to meet immediate and short-term requirements for food and agriculture, and remain available to meet possible longer-term needs. "Management of AnGR" is the combined set of actions by which a sample, or the whole, of an animal population is subjected to a process of genetic and/or environmental manipulation with the aim of sustaining, utilizing, restoring, enhancing and characterizing the quality and/or quantity of the AnGR and their products (food, fibre, draught animal power, etc.). From this definition, it is clear that "management" of AnGR encompasses all activities that ensure that the population is dynamic and is responsive to changes in the physical and sociocultural environment. "Management" also includes improving the understanding of the AnGR, i.e. characterizing and documenting the resources. Sustainable management includes those actions (including policy) that ensure that the AnGR meet present needs while also retaining their genetic integrity so as to be available for longer-term needs.
Whereas it is difficult to imagine an in situ conservation programme that does not involve local communities, "community-based management" emphasizes the involvement of communities, but does not necessarily rule out the use of ex situ approaches as complements to in situ options. Moreover, it is possible to have an in situ conservation programme in which a selected number of farmers are involved as individuals, e.g. through incentive schemes, but without collective community involvement. Of course, such programmes may have less probability of success than "community-based in situ AnGR management" programmes. A good example of a community-based programme for AnGR management with an ex situ component might be the open-nucleus breeding approach involving farmers and a nucleus herd, with or without use of modern reproductive biotechnologies such as artificial insemination and embryo transfer.
What is a breed in the current context?
The concept of a breed, in which all members have a pedigree tracing their ancestry, was developed primarily in western Europe during the eighteenth century. Today, in the developed world, breeds are recognized as distinct intraspecific groups, the members of which share particular characteristics that distinguish them from other such groups, and formal organizations usually exist for each breed or breed group. In its strictest sense, a breed designates a closed or partially closed population - mating pairs are drawn only from within the population and relationships among individuals are documented. Members of a breed have developed under the same selection pressures and share common ancestry. In reality, breeds are dynamic. To be successful they must constantly change in response to changes in societal needs as reflected in market demand. The change is achieved through selective breeding and "injection" of bloodlines from other breeds. However, for a population to retain its identity as a breed, there has to be less gene flow from outside relative to mating among members of the breed.
As has been pointed out, even in developing tropical regions of the world, where the term breed as a formal designation often has little meaning, there exist distinct strains or "types" that owe their identity to either geographical or cultural separation, the latter being the result of long-term artificial breeding by local communities. Livestock populations, developed in different sociocultural, ecological or geographical settings, will become genetically distinct as a result of genetic drift and differential selection pressures, natural and artificial, provided they have also been, to a considerable extent, reproductively isolated from other populations developed under different conditions. Thus, the indigenous livestock from different regions of the world should probably be assumed a priori to represent different "breeds". It seems clear that populations with different adaptation characteristics or possessing unique physiological characteristics should be recognized as different breeds. This distinction should be drawn even if the populations are shown to be relatively closely related based upon measures of genetic distance.
The World Watch List for Domestic Animal Diversity (WWL-DAD) prepared by FAO (2000) has defined a breed as: either a homogenous, subspecific group of domestic livestock with definable and identifiable external characteristics that enable it to be separated by visual appraisal from other similarly defined groups within the same species, or a homogenous group for which geographical separation from phenotypically similar groups has led to general acceptance of its separate identity.
When breed identity is documented through pedigree records, one can presumably document the time of genetic isolation and thereby place some boundaries on likely distinctiveness between candidate breeds. However, only a relatively small proportion of the worlds livestock is listed in herd-books. When potential "breeds" are physiologically similar and have overlapping and often large ranges, we should probably then utilize measures of genetic relatedness to help sort out breed distinctions. Thus, if we have basically similar animals across a wide area (e.g. fat-tailed sheep in Africa), with little phenotypic variation among populations and little reproductive isolation between adjacent populations, estimates of genetic distances among populations at the extremes of the range may be very helpful in assigning estimates of genetic uniqueness and, more important, in assigning conservation priorities relative to other populations. Where herd-books exist that appear to document genetic uniqueness among breeds, measures of genetic distance can supplement this information in situations where breeds appear otherwise quite similar.
Thus, we work at two levels. First there are the populations that are clearly distinct and unique, based on adaptations and physiology. There are degrees of "distinctiveness", but either a group of knowledgeable breeders and scientists or a good discriminate analysis should be able to pick these out. Next come the populations that are not so easily distinguished, and whose genetic uniqueness must be determined. It is at this level that "breed" distinctions become a challenge, requiring a combination of genetic and cultural distinctions. With or without genetic data, we should accept the cultural definition as being valid and try to gather information from the local communities that may explain whatever results we obtain from genetic studies. This is basically the approach taken by the AnGR programme of the International Livestock Research Institute (ILRI) to recognize breeds when the owners suggest that they are distinct (for example, through on-farm surveys), and to proceed to attempt to acquire objective measures of genetic relatedness. In making these distinctions among breeds and in accordance with the above analysis, especially when several apparently similar breeds are found in the same area, a population can be accorded a breed identity when groups of farmers in the area can be identified who: i) claim to be raising animals of a distinct type; ii) can reliably recognize that type; iii) exchange germplasm only with other breeders dedicated to holding animals of the same type; and iv) indicate that such breeding programmes have been going on for many generations.
The making of breeds in traditional systems
Social and cultural contexts of breeds or strains
In order to appreciate the role that traditional communities can play in conservation and management of AnGR, it is important to understand the sociocultural underpinnings of breed development. Social institutions and cultural traditions provide the context that determines the animal management choices available to farmers. Indeed, social and cultural forces are often the most important factors in diversifying livestock (and livestock production systems) and in developing breeds. Social and cultural factors influencing the decisions a farmer makes include traditional practices, local ways of life or the ethnic or community identity to which the farmer belongs. The value of a breed in the lifestyle or identity of a particular social group is what encourages its maintenance. Breeds may have specific, unique traits valued by the community that are not obtained from other "exotic" animal populations. Breeds may also be valued because of their place in local traditions - e.g. for their use in religious or other cultural ceremonies or because they provide products valued in traditional meals or medicinal practices requiring specific qualities. Rege, Aboagye and Tawah (1994) have described the unique role of the Muturu breed of cattle in southern Nigeria. Although the breed is not usually milked because the modest milk yield is just sufficient for the calf, milk is extracted by traditional doctors for medicinal preparations. The animals and their hides are used mainly for ritual sacrifices and ceremonies, particularly funerals. When a pagan dies, for example, one or more Muturu oxen are sacrificed and the corpse is rolled up in the hides of the slaughtered animals, while the meat forms part of the ceremonial feast. In Igboland, the Muturu were traditionally considered as sacred (juju cattle) and were the property of local deities or were dedicated to the shrine. Among the Koma people, the Muturu is kept as a semi-feral population. Individual animals needed for sacrifice have to be hunted down. Ticks are not removed as these are considered harmless to the Muturu. All these complex traditional practices and beliefs have important implications for the present-day genetic diversity in the Muturu.
Social organization and institutions in a community can influence farmers access to, and management of, household and community-level resources, affecting their action regarding the farm animal genetic diversity. For example, land tenure and ownership systems vary between and within communities in terms of private or communal ownership, equitability of distribution, size and number of parcels of household land, and intrahousehold access to land. A farmers landholdings and how they are distributed, their sizes and quality may influence decisions about breed choices and allocation of land area among breeds. In the central highlands of Kenya, for example, it was the introduction of individual ownership of land in the 1950s and 1960s and the accompanying fencing that promoted the successful replacement of indigenous livestock breeds with exotic dairy cattle germplasm, which required strict disease control and hence minimal animal movement. Today, even the "roadside grazing system" practised by the landless is characterized by cross-bred cattle.
Traditional breeding goals and objectives
In comparing the evolution of Western agriculture with that of developing countries, one often encounters the statement that the animal (and crop) genotypes (breeds, strains, landraces) currently found in developing regions are predominantly a result of natural selection, while those in developed countries are a product of "many generations of artificial selection". But this is a misconception. For centuries, farmers in the traditional sector everywhere in the world have used phenotypic features - physical characteristics, measures of yield, product quality, adaptive attributes, etc. to identify and select their breeds, strains or landraces. In livestock, these selection criteria may take a wide range of forms and are usually linked to the genetic diversity of the breed. They are used by farmers to distinguish and identify breeds/strains and are commonly the basis for farmers culling decisions. The Fulbe clan, the Udaen, after which the Uda sheep, a Sahelian breed of West Africa, is named, provides a good example. Individual Uda flock-owners go great lengths to maintain bloodlines based on certain desired traits.
Because these phenotypic characteristics are used to identify or distinguish breeds, they are often the basis for the names farmers give to specific animal types or strains, usually within a range of animals of a particular type or breed owned by the broader community. Thus, the large diversity of coat colour patterns in the Nguni cattle of southern Africa is classified by the Nguni herders into an elaborate system of names, each referring to a set of colour combinations. Phenotypic characteristics are also used in designating preferred or valued traits and as "criteria" for making selection decisions to achieve selection "goals".
Evidence for the existence of distinct animal types or breeds in traditional African society and links of these animal types to ethnic communities is provided in these statements in the writings of Charles Darwin (1868):
At the present day various travellers have noticed the differences in the breeds in Southern Africa. Sir Andrew Smith several years ago remarked to me that the cattle possessed by the different tribes of Caffres, though living near each other under the same latitude and in the same kind of country, yet differed, and he expressed much surprise at the fact. Mr. Anderson in his letter to me says that, though he will not venture to describe the differences between the breeds belonging to the many different sub-tribes, yet such certainly exist, as shown by the wonderful facility with which the natives discriminate them.
Breed as a unit of genetic diversity
Recognizing the names farmers give to animal populations is important because the "farmer-named population" is the unit that farmers manage and use as basis for selection decisions. The name or description of a population as used by the farmer may not only be related to physical characteristic(s), but could also relate to the original source of the breeding material. Both names and traits that define these names may also be related to the biological performance: egg production, size, shape, colour, milk yield or quality, aspects of adaptation, etc. Farmers perceive these attributes at various stages of the animals growth and development. Clearly, the set of traits that farmers use to identify a "breed" may be complex and are always deeply embedded in the culture and tradition of the community. To the extent that phenotypic characters are expressions of genetic characters at the level of transcribed genes or recessive alleles, phenotypic diversity (maintained by farmers) and genetic diversity can be seen as being equivalent views of the level at which diversity value resides. Any attempt aimed at "improving" or conserving the breed has to understand these complexities, which must be taken into account when developing the intervention strategies.
As has been pointed out above, breeding objectives are implemented at the breed (or equivalent) level. Formation of breeds is also aided by environmental factors. It is the differentiation of livestock species into breeds that has been responsible for the creation of populations adapted to specific environments and used for specific functions/products. Consequently, the breed represents the most important unit in genetic diversity assessment and conservation. For example, research to date suggests that between-breed variation accounts for about 50 percent of the total genetic variation in livestock species (FAO, 2000).
Consistency in names of breeds/strains
Farmers may or may not be consistent in naming and describing breeds or strains. It may happen that even within a village or community, different clans or families have different names for what is essentially the same breed or strain. As has been pointed out, this may be because of differences in valued traits, functions or other phenotypic characteristics or use of names linked to the origin of the germplasm, separately or in combination with valued characteristics. To the extent that these are important not only in understanding the evolutionary history of the genetic diversity in the breed, but also as an input in formulating management strategies that are relevant to the communities, it is crucial that any discrepancy in names be discussed and reasons for differences understood.
It is very interesting that, in Africa, at the broad (national and/or sub-regional) levels, present day "breed" names, assigned principally by scientists, tend to have geographical connotations or to be associated with names of tribes or ethnic communities. Examples include the Boran cattle, Mashona cattle, Nguni cattle, Gala goat and Somali sheep. Whereas this "naming system", principally an external construct, provides a useful analytical basis for broader environmental and cultural links to animal diversity, it oversimplifies the situation and ignores potentially important subtleties at local levels that could provide insights into the historical breeding systems that have shaped existing genetic diversity. Thus, any study aiming to understand breeds, as they exist today, must include on-farm surveys designed in such a way that the indigenous knowledge by local communities can be captured, analysed and subsequently used in designing AnGR management initiatives. To assume, for example, that the sheep distributed in the whole of coastal West and Central Africa and currently lumped together under one name - the Djallonké - are a single breed, would be to ignore the existence of local traditional animal-breeding culture across these countries and to suggest that, at some historical period, there existed a super-regional sheep-breeding programme! Granted, the natural environment of the subhumid and humid zones has exerted some (strong) selective pressure to produce sheep of similar characteristics. Nonetheless, the influence of local environments and, most important, the artificial breeding efforts of the diverse communities that own the "Djallonké" must be considered.
Other facets of community involvement
Agro-ecosystem influences on animal genetic diversity
Agro-ecosystems provide the ingredients for the evolution of animal genetic resources and the emergence of strains and breeds. This occurs through provision of stresses as well as opportunities. The responsible agro-ecosystem variables include such abiotic factors as temperature, relative humidity, rainfall and soils, and biotic factors such as parasites, pathogens and vegetation. As managers of these factors, the farmers are, as discussed above, an important biotic component of the system. Stochastic changes in these factors, temporally and spatially, from the micro-environmental to the ecological scale, do occur. These are the forces that drive the genetic changes, leading to the emergence of animal subpopulations that are adapted to the particular conditions of their immediate ecogeographic setting, and contributing to the formation of strains/breeds. An understanding of these interactions is crucial to the "on-farm conservation" of indigenous AnGR.
Community involvement in agro-ecosystem characterization
Understanding farmers systems of classification for the different aspects of their ecosystems may provide insights into the processes fostering conservation of animal diversity. In many ways, the farmers classification will coincide with the scientific ecological classification as they also principally use climatic and physical (e.g. landform and vegetation) variables. However, farmers may also classify ecological features based on their historical and cultural significance (Martin, 1995) and these may have important implications for the understanding of the overall system. Farmers ecological classification systems may serve as an indication, to the researcher or development agent, of which features are particularly important in the agro-ecosystem in relation to the range of animal breeds or strains being maintained and how these features might be incorporated in the design of sustainable programmes for the management of AnGR.
Programmes for the "development" of the livestock sector in Africa and, indeed, many developing regions of the world have, until recently, focused on the introduction of exotic livestock breeds. In the absence of objective and credible data at production-system level, impressionist accounts of indigenous breeds suggested to development agencies, including government agricultural extension, that indigenous breeds had low productivity compared with European breeds. Little attention was given to the economics of production under the harsh conditions and minimal inputs. Moreover, occasional "successful" farms or government stations evaluated on the basis of short-term performance, and usually based on (often subsidized) economically unrealistic management practices, seemed to validate such strategies (Dunbar, 1970). Save for a few exceptions - high-input farms in Kenya and in some countries in southern Africa, supplying urban markets with relatively developed infrastructure - imported breeds have largely been a failure in Africa. Fortunately, there now seems to be a realization of the need to refocus attention, not only on indigenous breeds, but also on indigenous knowledge.
From a livelihoods perspective, it can be said that the sectoral approaches to agricultural development have, to date, generally ignored the substantial natural and social capital accumulated over millennia by traditional livestock keepers (Blench, 1999). The conclusion that development should draw on this capital is only now slowly being reached. Unfortunately, while this realization has led to a paradigm shift among technicians and national planners at some levels in many countries, there is still an urgent need for considerably more knowledge about indigenous livestock breeds, their ecological adaptations and their productivity under traditional management. Some work has been initiated at national, subregional and continental levels and information is slowly being accumulated, principally on cattle and, to some extent, sheep and goats. The situation is much more dismal for indigenous pigs, camels, chickens and other poultry species. Be it research to improve understanding of indigenous breeds, or initiatives to develop programmes for their sustainable management, community involvement is the route to success. In crop agriculture, participatory plant breeding is now generally accepted and widely applied in numerous developing countries. Livestock development remains primarily driven by imported technological packages (e.g. artificial insemination, exotic germplasm) and very limited involvement of communities in their implementation. This has to change.
Blackburn, H., de Haan, C. & Steinfeld, H. 1996. Livestock production systems and the management of domestic animal biodiversity. In J.P. Srivastava, N.J.H. Smith & D.A. Forno, eds. Biodiversity and agricultural intensification, pp. 95-106. Washington, DC, the World Bank.
Blench, R. 1999. Traditional livestock breeds: geographical distribution and dynamics in relation to the ecology of West Africa. Working Paper 122. London, Overseas Development Institute.
Brush, S.B. 1991. A farmer-based approach to conserving crop germplasm. Economic Botany, 45: 153-165.
Darwin, C. 1868. The variation of animals and plants under domestication. Popular Edition 1905. London, John Murray.
Dunbar, G.S. 1970. African Ranches Limited, 1914-1931: an ill-fated stock-raising enterprise in Northern Nigeria. Annals Assoc. of American Geographers, 60: 102-123.
FAO. 1992. The management of global animal genetic resources. FAO Animal Production and Health Paper No. 104. Rome. 309 pp.
FAO. 2000. World watch list for domestic animal diversity. Third edition. Rome. 726 pp.
IUCN/UNEP/WWF. 1991. Caring for the Earth: A strategy for sustainable living. Gland, Switzerland.
Martin, G.J. 1995. Ethnobotany. People and Plants Conservation Manual Series. London, Chapman and Hall.
Rege, J.E.O., Aboagye, G.S. & Tawah, C.L. 1994. Shorthorn cattle of West and Central Africa II. Ecological settings, utility, management and production systems. World Animal Review, 78: 14-21.