Simulações de sustentabilidade e de formação de expectativas na produção de alimentos na agricultura familiar do Maranhão

Abstract

The temporal and spatial distribution of rainfall patterns in Maranhão State aims to be observed at least ten (10) distinct regions. In this State the crops of rice, beans, cassava and maize grown, in general, without the use of more accurate technologies, can constitute important thermometers to measure sustainability and to extract expectations of family farming in the state. Thus, this research aims to: a - Evaluate, experimentally, the possibility of maintaining aggregate production of rice, beans, cassava and corn without expanding the cultivated areas, from the historical average in each of the ten regions where Maranhao State was subdivided according to their weather conditions between 1980 and 2016; b - Assess the ways farmers of Maranhão State, growers of rice, beans, cassava and corn form production expectations. Secondary data were extracted from the Municipal Agricultural Research and the National Institute of Meteorology. The work draws two methodologies. For the application the first method, the the development of two grupos: the experimental and the controle. In the experimental group is simulated the areas of these cultures are maintained constant in their historical average. With this procedure, production in above the average areas effectively reaches the levels observed during the reporting period by increasing their productivity (yield pre hectare). In the control group areas is maintained as were observed. Thus, at this stage the research simulates how should increase crop yields in order to compensate the non-expansion of areas. By the experimental procedure, to test if crop yields gain in economic, compared to the way farmers preceded throughout the series, it was created the sustainability index (ISUS) by using linear programming techniques. This was the instrument used to test for statistical differences between the two groups evaluated in the survey. The second part of the work simulate up the ways in which, probably, farmers format their gross income expectations. For this part it was introduced the randomness factor which is the annual precipitation of rainfall. To generate the probable ways in which farmers format their gross income expectations are used time series analysis techniques. It was tested autoregressive models, moving averages, and auto regressive moving average simultaneously. The results showed that in all ten regions is possible to produce rice, beans, cassava and corn in a sustainable way without expansion of areas from a certain value: its historical average. To do this the farmers must have to access technologies that enable higher levels of productivity. Also, it was found that the Maranhão State farmers formulate expectations based on information from the immediately previous years, given that all the adjusted models are fist order auto regressive.