Abstract:
Principal component analysis was used to identify the factors explaining the most variability between the farms pooled across districts. The identified factors were then used to run hierarchical clustering of the farms to derive farm types that were relatively similar with respect to the principal factors accounting for between farm variability. The differences in the principal factors between farm typologies were evaluated to deduce the general livelihood strategies adopted within each farm typology. Farm household characteristics were evaluated by comparing means of the variables representing household human, natural, physical and financial capital, and how they are related to the major livelihood strategies.