Dispersed settlement in rural Ethiopia coupled poor infrastructure including roads, transportation and telecommunication is one of the reasons for the low efficiency of AI service. Efficient delivery of AI service to individual farmers as and when each cow shows oestrous signs is impracticable. The AI service and delivery of improved dairy genetics to smallholders could be improved by introducing hormonal oestrus synchronization and mass artificial insemination . The effort to improve the dairy cattle crossbreeding/AI program through hormonaloestrus synchronization has been below expectation due to factors related to inefficient AI service, poor infrastructure and farmers’ awareness, although the synchronization technology has been successfully demonstrated through action research .An important characteristic of the urban and peri-urban dairy farming systems in and around the regional towns identified in the current study is that local dairy cows are numerically equally important in all the three systems, though this varied between the geographic regions studied.

On the contrary, studies on the characterization of the smallholder dairy systems have considered keeping local cows as characteristics of the rural system and not of the urban/peri-urban systems. Local breeds have received little attention from the national dairy cattle genetic improvement programs which has been taken virtually as synonymous to crossbreeding programs. Yet after decades of implementation of the national Holstein Friesian dairy cattle crossbreeding program, the proportion of the crossbred animals has not exceeded 1% of the national herd in Ethiopia .Thus the local dairy breeds are still prominent in the dairy genetics landscape of developing countries and thence continue to be an important resource in genetic improvement and dairy development strategies in all dairy farming systems excluding the systems located in and around the big cities. Besides their numerical importance, the indigenous cattle breeds are generally characterized as multipurpose animals, can be managed in low input production system and are adapted to marginal environments.

However, the current data on herd genetic structure with admixture of different breeds and genotypes including in the rural system which is believed to be the sanctuary for the indigenous genetic resources could be an indication of absence of well-planned and designed crossbreeding program, which could lead to indiscriminate crossing and threat to the adapted indigenous resources. The threat to the adapted indigenous resources is however more pronounced in the peri-urban/urban system according to the current data and elsewhere in Africa where, for instance, the proportion of crossbreds reached close to 40% in 2015 from 9% in 2004 in peri-urban system in Bamako. However, the local breeds are inherently low milk producers as they have been naturally selected for adaptive traits and not functional traits . Selective breeding could be the best option to improve the genetic merits of the indigenous breeds. However, to implement effective selection programs in village herds,there are constraints relating to the characteristics of smallholder village herd structure, infrastructures and attitudes of the public and private sectors towards selective breeding programs.

For instance in the current study local breeds are more important in West Gojjam zone than in West Sho zone which is located in the greater Addis Ababa milk shed where there is better access to breeding services and markets for fresh milk.Of the total 360 farms surveyed in West Shoa and West Gojam zones in this study, only 50.6% and 38.7% of the farms, respectively, conformed to the typical breeding characteristics identified by the analysis model for rural, peri-urban and urban dairy farming systems. This has an important implication in the designing and introduction of appropriate dairy development interventions. This is particularly relevant to the peri-urban system where the highest misclassification was observed. Identification of determinants of herd genetic structure regardless of the influence of farming systems showed that access to breeding services and land resources were the most important factors. The low proportion of crossbred cows in existing herds in the current study sites is contrary to farmers’ preferences for crossbred cows as elicited for the same study site and elsewhere in the highlands of Ethiopia.

The major constraint to meet farmers’ preferred dairy breeds and genotypes is the limitation in the AI service. The conception rate to first AI service has been reported to be as low as 27.1% , and the major challenge as reported by farmers is poor heat detection where AI is accessible, whereas access to reliable AI service has also been reported to be very low even in the peri-urban areas in an extensive study in the four highland regions of Ethiopia . The service is neither better around Addis Ababa where dissatisfaction has been expressed by about 46.7% of farmers surveyed . It has been argued that utilization and improvement of the desired crossbred population can only be efficient in situations where breeding programs with well-defined breeding objectives, breeding structure and infrastructure are developed,which is often lacking at smallholder level in the tropics.