Genomic breeding in simple terms refers to the inclusion of deoxyribonucleic acid (DNA) or genomic information to select superior animals and make them parents of the next generation (Meuwissen et al., 2001, 2016). Thus, genomic selection simply is application of the knowledge of genetic variations found in the genome and their relationship with traits or phenotypes (e.g., milk yield, body weight, egg size, etc.) in selection for improved productivity (e.g., litter size, milk yield, etc.). The application of quantitative genetic theories, statistical approaches, artificial insemination and organized breeding practices resulted to rapid gains in livestock traits in the last nine decades (Blasco, 2013; Hill, 2014; Oldenbroek and Waaij, 2014; Weller, 2016). Mostly, the exact mechanisms behind these gains were not known but with the discovery of the DNA structure and developments in DNA sequencing and genotyping techniques, knowledge on the association between DNA variations and livestock traits began to emerge. Thus, with increasing demands for animal products by an ever growing population and changing societal needs, the animal breeding act needed to evolve to incorporate genomic information in order to speed up response and increase productivity.

Genomic breeding started with the application of marker assisted selection considering a few markers at a time and has evolved to the use of thousands of markers and even whole genome data (Hayes and Goddard, 2001; Hayes et al., 2013). Genomic selection entails the estimation of breeding values from markers spanning the entire genome. The estimation of marker effects is carried out within a reference population (a population of individuals with phenotype and marker genotype information). These effects are then applied to select candidates with marker genotype information without phenotypes to estimate genomic breeding value (GEBV). The reliability and accuracy of this approach depends on many factors including the number of individuals genotyped, the density of the markers on the genome, effective population size, the genetic relationship between the reference and predicted populations, the nature of the traits and the applied methods, etc. (Habier et al., 2007, 2011; Bolormaa et al., 2013a,b, 2014; Meuwissen et al., 2016). Genomic selection for milk and beef traits has been successfully implemented in several countries including Australia, Canada, France, Germany, Great Britain, Ireland, the Netherlands, New Zealand, the Scandinavian countries and the United States of America (Silva et al. (2014); Weller et al., 2017). Genomic breeding application in these countries is facilitated by many factors including: (1) large population of animals; (2) specialized farms; (3) comprehensive data on animals; (4) access to genotyping platforms which makes single nucleotide polymorphism (SNP) genotyping more cost effective; (5) available resources (finance, technical knowhow) to accomplish genotyping; (6) existence of breed associations; (7) application of artificial insemination; (8) development of breeds for specific purposes; (9) large scale international breeding companies that sell semen from high performing males for use in breeding for specific traits; (10) creation of farmer organizations, (11) implementation of national evaluation schemes; (12) development of statistical models to handle large data and (13) computing infrastructure to deliver genomics information which helps to facilitate genetic gains in livestock.

In addition, several initiatives have been undertaken to make available data on all sources of genomic variation in livestock genomes to further increase the success of genomic breeding. Such efforts include, but not limited to, the 1,000 bull genome project with aim to re-sequence the whole genomes of 1,000 bulls and has already made available about 84 million SNPs and 2.5 million small insertions/deletions (Hayes and Daetwyler, 2019) and the international consortium for Functional Annotation of Animal Genomes (FAANG²) established to provide the infrastructure to detect and proficiently analyze genome wide functional regulatory elements (DNA methylation, histone modifications, chromatin remodeling, non-coding RNA) in animal genomes (cattle, chicken, goat, pig, and sheep) necessary to understand how variation in gene sequences and functional components determines phenotypic diversity, and how this is translated into complex phenotypes; and thus fill the genotype-to-phenotype gap that is missing in current livestock improvement programs (Andersson et al., 2015; Tuggle et al., 2016). Major gains achieved with the use of genomic information and implementation of genomic selection include higher rate of genetic gain, increased reliability of predicting breeding values, higher intensity of selection, shortened generation interval, selection of animals possible at early age, and rapid genetic improvement in lowly heritable traits (e.g., fertility, lifespan, health, etc.) (reviewed by Hayes et al., 2009; Ibeagha-Awemu and Khatib, 2017; Weller et al., 2017; Mrode et al., 2018).

The application of genomic selection in Western countries and the advances that have been made in breeding (e.g., dairy traits) have been driven by the economic needs of the producers. However, challenges regarding sustainability of livestock production necessitate consideration of the economic, societal and environmental factors. A focus on increased milk production for example and intensive selection for this trait for several decades resulted in a deterioration of many traits like fertility, udder width/circumference and disease resistance (e.g., mastitis, metabolic diseases) traits, etc., and an increase in its ecological footprint (e.g., greenhouse gas emission) (Boichard and Brochard, 2012; Egger-Danner et al., 2015). These factors together with growing demand by consumers for animal safety warrant that successful programs for sustainable animal improvement should create a balance between selection for traits of economic value, animal health, conformation traits, adaptation traits, animal welfare and environmental foot-print.

The successes of genomic selection in Western countries mentioned above were possible through organized and sustained breeding practices supported by government regulations, finance and involvement of private companies. The picture for the majority of African countries is different given that, most livestock are kept for multi-purposes (meat, milk, traction, hides/wool, as a savings account, social status, cultural reasons, etc.), in small herds and flock sizes, under small scale to mid-scale low performing and low input systems, and lack of enabling government policies and financial support. Thus, procedures to increase livestock productivity in Africa in the era of genomic breeding must take into consideration the different production systems, ecological zones and participation of all stakeholders.

In majority of African countries, livestock production is managed under small to large scale systems (Table 3). Small scale production systems include pastoral, agro-pastoral and mixed smallholder farming. Large scale systems include ranching, large scale commercial farming, cooperative farming and state owned farms. About 70% of livestock productivity occurs under the small scale systems characterized by small animal population sizes, low inputs and outputs, etc. Devising appropriate policies for such systems with the right government support is of utmost importance in increasing livestock productivity for food production and income generation. The characteristics of the various systems of livestock production are summarized in Table 3. Under predominantly small scale farming systems, it is important to determine whether or not such systems are ready for genomic breeding. Genomic breeding implementation relies on available genetic resources/diversity, and genomic variation and its association with desired traits.

As discussed above, the African continent is home to diverse livestock populations which also display rich genomic variations within and between breeds and have acquired special adaptive characteristics that support adaptation to poor quality feed, limited water supply, hot environments and disease (Psifidi et al., 2016; Mrode et al., 2018; Lawal et al., 2018). Some of these characteristics are summarized in Table 4. The indigenous breeds have acquired important characteristics for survival in their environments and in addition to being developed systematically, should be conserved for future survival and exploitation. Therefore, conservation of local breeds (highly utilized and less utilized) must be part of national breeding plans and should not be an exercise undertaken by individual farmers. Systematic breed development strategies including selection within breeds, controlled crossbreeding and upgrading programs and development of new breeds to exploit special adaptive and/or production traits must be done under organized systems with specific goals. This will address the practice of indiscriminate crossbreeding between local breeds and, between imported breeds and local breeds that is eroding the continent’s animal genetic diversity, a much needed resource for present and future exploitation.

There is still a lack of understanding about situating livestock development programs within prevailing low-input production systems, societal preferences and environmental conditions. Most international development programs have been based on ‘top-down approaches,’ considering single commodities and technology focused orientation with little or no participation of farmers nor formation of strong farmer-based institutions (Kaasschieter et al., 1992) and with no regard to prevailing environmental conditions. Majority of livestock are produced under the small holder system composed of pastoralists, agro-pastoralist and small holder farm families which have different breed preferences, inputs and challenges. Therefore, the prevailing production systems and livestock production potential under the different systems must be characterized and their specific needs identified in relation to local preferences and market needs, with the participation of producers. Data on productivity of local breeds under their prevailing conditions of production are largely unavailable. Such data is necessary as it will form the basis for improvement plans for each system. Indigenous African breeds are generally considered as underproductive without giving thought to the low-input and harsh environments in which they are raised. For example, under farmer management, the Butana and Kenana zebu breeds of Sudan produce averagely 538.26 and 598.73 kg of milk per lactation, respectively, while under research station conditions, they produce 1,400–2,100 kg of milk per lactation, respectively (Musa et al., 2005, 2006; Yousif and El- Moula, 2006). This implies that, although Butana and Kenana seem to produce less under low-input systems, they have the ability to produce more given improved conditions of nutrition, health care and production management. Another factor is the bias in judging local breeds based on parameters that have been selected and developed in exotic breeds. An example is the focus on lean specialized pig breeds like Large White at the expense of local relatively fat breeds that are well adapted to local environmental conditions. As such, policies are put in place that disfavor production of local pigs but encourage their replacement with exotic breeds (Chimonyo and Dzama, 2007).

Furthermore, emphasis has been placed on realizing quick gains by adopting ‘shortcut approaches’ like introduction of exotic genes that have been developed over a long period of time in different environments. This exotic stock does experience genotype by environment interaction which dramatically lowers performance in the local environment where they are introduced. African livestock have acquired the characteristics necessary to produce under their prevailing environmental conditions and on minimal resources. Before introducing exotic genes in a controlled manner, firstly, the productivity of local breeds must be assessed under optimal conditions (e.g., adequate feeding, housing, and disease management). The main limiting factors of local breed productivity could just be management and limited feed resources and disease control measures, as exemplified in Butana and Kenana cattle (Musa et al., 2005, 2006; Yousif and El- Moula, 2006). Under optimal management conditions, local breeds could be selected for desired traits in their prevailing environmental conditions. The adoption of most ‘shortcut approaches’ utilizing exotic genes has generally not resulted in substantial gains and sustainable long-term increases in productivity or contributed to poverty alleviation. Most donors or policy makers are only interested in immediate short term gains (visible) with the result of reckless crossbreeding of indigenous cattle resources with exotic breeds or their complete replacement with exotic breeds. These ‘short term gains’ are usually lost when such programs end and usually, the offspring of crossbred animals underperform under the prevailing conditions or lose the adaptive productive ability of the local breeds.

Setting a clear breeding goal is a prerequisite for animal improvement planning and implementation of genomic selection. A definition for animal breeding goals planning as a procedure with ethical priorities and weighing of market and non-market values has been suggested (Olesen et al., 2000). The decision to develop an animal for specific products has primarily been for the common interest of the farmer or the society or market demand. For example, increased market demand for milk and its products drove dairy breeding objectives in Western countries toward increased milk yield which unfortunately has resulted to problems of fertility and huge environmental footprints (Gill et al., 2010; Wall et al., 2010; Hayes et al., 2013; Knapp et al., 2014). Hence, sustainable animal breeding goals should consider market economic and non-market value traits, farmer specific needs, social, ecological and environmental needs supported by appropriate government policies, education, more cooperation between stakeholders and, short and long term needs, etc. The breeding goals must be adapted to fit each production system and environment. In recent times, breeding goals considering new phenotypes or non-traditional or production oriented traits and genetic traits of relevance to breeding sustainability have been proposed for cattle, sheep and pigs (Merks et al., 2012; Banga et al., 2014; Miglior et al., 2017; Molotsi et al., 2017). Recently, results of a survey of 160 farmers in southern Mali identified draft power and savings as the most important production objectives while preferred traits included fertility, draft ability and milk yield, in that order (Traoré et al., 2017).

Optimal animal productivity is supported by adequate nutrition and disease management. Options for quantitative and qualitative improvements of the feed resources according to the needs of the different production systems are required for sustainable livestock systems (Duncan et al., 2013; Thornton and Herrero, 2015). For example, under the pastoral system, communal access to rotational grazing pastures and fodder banks which should be maintained to ensure quality of feed resources will support sustained livestock production. Legislations instituting the development of watersheds, restrictions on indiscriminate burning of grazing land and use of such land for other purposes are of necessity. Other vital aspects include development of improved pastures and fodders, increased grain production, development of agricultural bi-products as feed resources, access to water resources, etc. Although some of the local breeds have adapted to the disease burdens of their environments, disease is still a major limiting factor to livestock productivity in the region (De Garine-Wichatitsky et al., 2013; Okuni, 2013; Vanderburg et al., 2014). Particular attention should be paid to disease control measures like access to drugs, vaccines and veterinarians, and sound management practices developed for each system (Maclachlan and Mayo, 2013; Miguel et al., 2013).

Precise phenotypic data is crucial for genetic improvement. In Western countries, systems have been put in place to support high throughput phenotyping (e.g., milk yield, milk component yields, feed intake, etc.) thus enabling the accurate and consistent collection of large amounts of data on animal productivity. The formation of livestock trade databases is worthy of consideration since livestock movement contributes to the spread of animal and zoonotic diseases. In Western countries, this database is important for researchers to describe mobility patterns, optimize disease surveillance and control and predict possible epidemic scenarios (Apolloni et al., 2018). Therefore, it is necessary to sensitize producers on the importance of data collection and record keeping on the productivity of their animals, as well as formation of data storage facilities that can facilitate data storage and sharing within and between countries.

Besides the common issues with infrastructure for general development of the economy, infrastructural development to promote livestock production within the continent must be considered such as basic equipment for sample storage, data collection and data trace, livestock markets, slaughter facilities, animal housing and pasture development. With advances in genomics and other omics technologies, the livestock sector in several countries has moved to the area of big data research and application (VanderWaal et al., 2017; Morota et al., 2018; White et al., 2018). African livestock infrastructure must be developed to optimize the use of big data. Moreover, farmers need to be sensitized and prepared for adoption of these technologies. Technologies need to be adapted to farmers’ specific needs according to the system of production since there are differences in farmers’ access to farm resources, technological inputs and differences in access to output markets (Birhanu et al., 2017; Feder and Savastano, 2017). For instance, adapted dairy technologies varies widely among smallholders (Staal et al., 2002; Abdulai and Huffman, 2005; Amlaku et al., 2012) and also strongly affected by their social networks (Amlaku et al., 2012). The environmental impact is now a major concern for livestock production the world over due to its impact on greenhouse gas emissions and consequently climate change. The livestock sector in Africa also pose a challenge to the environment and climate change depending on the management and farming system. For instance, the semi-arid region is faced with the problem of overgrazing of rangelands which is caused by population pressure and a decline in traditional management systems (Fratkin, 2001; Ngongoni et al., 2006).

The success of sustainable livestock development in any country or region hinges on development of national and or regional policies or guiding principles in the conduct of affairs. The decision by an international support organization or by a farmer group to import specific germplasm for crossbreeding with local breeds must be backed by national polices and priorities. Supply and demand policies favoring local production and supply chains will stimulate local production. Recognizing the important contribution of livestock production to the livelihood of farm families and to the nutrition and economy of the state/country necessitates a political commitment to stimulate, develop and financially sustain livestock development. The African Union has in place a Livestock Development Strategy (LiDeSa) for Africa (2015–2035) which was developed through an inclusive consultation process involving experts and stakeholders at national, regional, and continental levels (African Union Inter-African Bureau for Animal Resources AU-IBAR [AU-IBAR], 2015b). The strategy recognizes the central role played by livestock as a livelihood sustainer for rural Africa and with the support of a grant from the Bill and Melinda Gates foundation, seeks to transform the livestock sector by invigorating its untapped potentials. This is a laudable process that if implemented could truly transform lives. However, country level initiatives must follow suit for a transformed livestock sector to emerge in the continent. Today, the South African government is the only African government that is playing an active role in the conservation of animal genetic resources (Nyamushamba et al., 2017) and in supporting livestock breeding programs (Van Marle-Köster et al., 2017). Moreover, it is also important to take into account farmer’s preferences in the development of breeding polices (Wale and Yalew, 2007). Development of national policies and regional priorities should also focus on mitigation in the livestock sector due to the impact of climate change which varies with location. A program called climate-smart agriculture (CSA) has been implemented recently in the West African region and sub-Saharan Africa in general (Amole and Ayantunde, 2016). CSA is an approach that provides a conceptual basis for assessing the effectiveness of agricultural practice change to support food security under changing climatic conditions (Amole and Ayantunde, 2016).

Appropriate economic incentives are important for livestock genetic improvement. Breeding programs should be market-oriented and the government should provide the right incentives. Several countries have made efforts on the extension of market access as well as to encourage foreign trade. For example, the Ethiopian government has completely strategized to encourage foreign trade for sheep and goat products which has led to the creation of employment opportunities for its citizens (Nwogwugwu et al., 2018) or the emergence of livestock feed market in Ghana (Konlan et al., 2018).

Education and training, and information sharing are vital aspects in sustainable livestock improvement breeding. The training curriculum in higher institutions should be adapted to fully address the needs of the various production systems. Formal training of students and informal training of the producers is vital. Greater cooperation between universities, research organizations (international, national, and regional), producer groups, non-governmental organizations and governments will ensure the flow and sharing of information and knowledge (Figure 2). The International Livestock Research Institute (ILRI) in Kenya has and continues to train students, research professionals and farm groups in various aspects of livestock breeding and production and molecular biology/genomics techniques. ILRI’s work in consultation with and tailored to meet the needs of farm families has resulted to initiatives like the Dairy Genetics East Africa project (DGEA), African Dairy Genetic Gains (ADGG), etc³. These programs were tailored to increase farmer productivity and profitability through the use of cross-bred animal types supported by extension and training systems tailored to their needs. The influence of ILRI amidst other successes led to rapid increase in cow milk production between 2011 and 2012 in the East African region (Figure 1). A national milk recording scheme has been instituted and supported by the government of Kenya⁴. The challenges faced by the program include limited number of breed inspectors, unawareness by many farmers of the importance of livestock registration, delay in issuance of livestock certificates and poor record keeping by farmers. Some of the suggested solutions include: training of more livestock inspectors by breed societies in conjunction with government, create farmer awareness using sensitization campaigns through mass media, exhibitions, shows, field days and direct consultations with interested farm groups, decentralization of services and investment in manpower and infrastructure. National animal production research institutes and universities in the various countries can emulate some of the practices of ILRI given that farmer’s participation in the development of projects tailored to their needs is a vital aspect in the successes of such programs.

Regional and continent wide sharing of information is vital for the sustainability of the livestock sector. The Forum for Agricultural Research in Africa^5,6, a technical arm of the African union, coordinates and advocates for agricultural research-for-development in the continent. Regular meetings of stake holders (professionals, farmers, students, and industry) interested in the act of animal production in forums like the All African Conference on Animal Agriculture, country and regional conferences on animal production all serve vital roles in the flow of information and technology advancements. However, producer focused meetings that provide informal training to farmers are generally lacking.