This study characterized and classified the dairy production systems of the Bulbul river basin according to management, production and local knowledge aspects. A total of 66 producers were interviewed by means of a semi structured survey. Information was analyzed by three multivariable statistical techniques (analysis of principal components, analysis of multiple correspondences and cluster analysis). Farms were classified into three groups. Group 1 included farms with the lowest production level, less infrastructure, lower sanitary level, less supplementation practices, and lower education levels and technical advice. Group 2 included farms with higher production levels than group 1, higher proportion of improved pastures than groups 1 and 3, low use of feed supplements and low sanitary levels. Group 3 included farms with production levels significantly higher than farms in groups 1 and 2, better feed supplementation practices and sanitary levels, together with better infrastructure, higher education level of the owner, and better access to technical advice. Besides, farms in group 3 presented more knowledge and application of silvopastoral techniques. The study shows clear trends and relationships between management practices and productivity at the farm level; this knowledge can be used in designing strategies for farm technical advice.
