| The relationships between farmer's allocation of farm area to crops and the farmer's perception of anticipated modal prices and yields, and of their variability, were evaluated. It is hypothesized that the fraction of farm area allocated to a crop is: (1) positively related to its anticipated yield and price, and (2) negatively related to their variability.;Agriculture in Margarita was found to be very unstable due to ethnic and social conflicts, pest and diseases, recurrent natural disasters (especially major river floods), and failures in the export markets. It was also labor-intensive, capital-limited, strongly linked to export markets (subsistence agriculture is not practiced to a self sufficient extent), and agrochemicals (specially fertilizers) are not fully used.;This study showed that farmers used different decision rules to allocate farm area to different crops. In general, however, land was allocated to most crops (plantain Musa AAB, yam Dioscorea alata, ayote Cucurbita maxima, and maize Zea mays) in inverse relation to the farmer's anticipated variability in yields.;Diversification was linked to larger farms. A second (or third) crop is introduced only after insuring a certain amount of income from plantains, a reliable crop with acceptable returns. Polycrops performed well in large farms and with variable prices. Combinations of plantain with either (or both) yam, or ayote outperformed any other polycrop.;Plantain cultivation differed between diverse and monocrop farms. In 1991, diverse farms incurred lower production costs and yet expected higher plantain yields. In accordance with hypotheses of this study, farmers in diverse farms expected larger variability in price and yields for plantain. |
