Development of the livestock industry and its role in poverty alleviation in developing countries depends on how adaptive the production systems are to changing global environmental and economic trends. This paper characterizes dairy production systems in India, Tanzania, Kenya and Nicaragua, and describes the genetic and breeding technologies that hold promise for the advancement of global development goals. The dairy value chain has been prioritized for development under the CGIAR research programme on Livestock and Fish in Tanzania (East Africa), India (South Asia) and Nicaragua (Latin America), while ILRI is involved in research on dairy development in Kenya. In all the countries, a large number of smallholder farmers operating mixed crop–livestock production systems play a significant role in dairy production. In Tanzania, Kenya and Nicaragua, milk is predominantly produced by cattle of genotypes that differ both across countries and among production systems within the same country. In India, buffaloes contribute to a larger proportion of the national milk than cattle. Information on productivity per animal and on optimal genotypes to utilize within the smallholder production systems of all the countries is however limited. Crossbreeding and artificial insemination were identified as the most widely utilized breeding and reproductive technologies. Only in Kenya is there a national organization conducting livestock recording and monitoring productivity, however, the proportion of the dairy cattle population enrolled in the recording system is small (<2.5 percent). In all the countries, enhanced and adequately planned use of breeding and reproductive technologies, complemented with the relevant infrastructure, is needed to sustainably increase dairy productivity. The capacities of actors in the dairy value chain need to be developed in order to properly implement and manage improvements.

Africa has a shortage of animal products but increasing demand because of population growth, urbanisation and changing consumer patterns. Attempts to boost livestock production through the use of breeding technologies such as artificial insemination (AI) have been failing in many countries because costs have escalated and success rates have been relatively low. One example is Kenya, a country with a relatively large number of cows and a dairy industry model relevant to neighbouring countries. There, an innovative dairy marketing approach (farmer-owned collective marketing systems called dairy hubs) has been implemented to enhance access to dairy markets and dairy-related services, including breeding services such as AI. So far, the rate of participation in these dairy hubs has been slow and mixed. In order to understand this phenomenon better and to inform dairy-related development activities by the Kenyan government, we investigated which characteristics of AI services, offered through the dairy hubs, farmers prefer. To do so, we applied a choice experiment (CE), a non-market valuation technique, which allowed us to identify farmers’ preferences for desired characteristics should more dairy hubs be installed in the future. This is the first study to use a CE to evaluate breeding services in Kenya and the results can complement findings of studies of breeding objectives and selection criteria. The results of the CE reveal that dairy farmers prefer to have AI services offered rather than having no service. Farmers prefer AI services to be available at dairy hubs rather than provided by private agents not affiliated to the hubs, to have follow-up services for pregnancy detections, and to use sexed semen rather than conventional semen. Farmers would further like some flexibility in payment systems which include input credit, and are willing to share the costs of any AI repeats that may need to occur. These results provide evidence of a positive attitude to AI services provided through the hubs, which could mean that AI uptake would improve if service characteristics are improved to match farmer preferences. The dairy hubs concept is currently in the implementation phase with most hubs at startup phase, hence understanding which AI service characteristics farmers prefer can inform the design of high-quality and cost-effective AI services in the future.