Use of agent – based models in characterizing farm types and evolvement in smallholder dairy systems

Abstract

The ever-increasing demand for milk and dairy products has attracted research interventions on how milk yield can be increased for the context of smallholder farmers. While bearing significant contribution on milk production and fulfilment of the market demand, the smallholder dairy farmers are faced with challenges that hinder productivity. Among the challenges is the inadequate characterization of the dairy production systems and lack of knowledge on factors attributing to their growth. This has resulted in aggregation of the smallholder dairy farmers and lack of interventions tailored to suit particular farm types. By using Tanzania and Ethiopia as case studies, this research identified the main determinants for evolvement of smallholder dairy farmers. Evolvement in this research refers to, gradual increase in milk yield. The factors that determine evolvement for individual farm typologies were identified by using cluster and frequent pattern analysis. The differential influence of the identified determinants towards increase in milk yield was studied by using Agent-based modelling and simulation where each factor was observed. Six farm types were identified for Tanzania and four for Ethiopia. The characteristics of the farm types were enriched by frequent pattern analysis with confidence level 60% - 97%. Agentbased modelling revealed that, income and farm-based determinants influenced an increase of

up to 7.58 litres above the average (13.62 ± 4.47) for Ethiopia. For Tanzania, farm and farmerbased determinants influenced an increase of up to 7.72 litres of milk above the average (12.7

± 4.89). The identified determinants could predict up to 96% and 93% of the variances in milk

yield for Tanzania and Ethiopia, respectively.

There was an increase in milk yield based on the identified evolvement determinants; from baseline data average milk yield of 12.7 ± 4.89 and 13.62 ± 4.47 to simulated milk yield average of 17.57 ± 0.72 and 20.34 ± 1.16 for Tanzania and Ethiopia, respectively. Dairy development agencies should consider the disaggregation of dairy farmers and prioritization of the determinants identified in this research for evolvement of dairy farms. In future, it is important to develop a web or mobile application that can inform smallholder dairy farmers about the identified evolvement determinants to aid on-farm decision making.