Melhoria da precisão nas estimativas de produção em culturas olerícolas relacionado ao número de plantas e colheitas

Abstract

The variability in crop production oleraceous affects planning and the quality of experiments, inflating the experimental error, and leading to erroneous conclusions with low experimental precision and reliability. The objectives of this study were to scale the number of plants and crops necessary for there to be improving the precision of estimates of production in vegetable crops, test geostatistics and modeling tool and verify the spatial dependence of fruit production of vegetable crops. From the blank experiments production data, obtained from protected cultivation with crops of zucchini, peppers and snap beans, the Department of Agronomy of the Federal University of Santa Maria, was applied the response surface model with variable dependent on the coefficient of variation and independent plot size and number of samples. Semivariograms were made in production data processed in indicators with the average cutoff point, in different simulations of plot sizes and crop groupings, and set the theoretical models spherical, exponential and Gaussian. For the zucchini the best combination of plot size and the number of crops, is seven base units and half of the production cycle, in both growing seasons. For the peppers, using 23 basic units and three harvests per plot grouped, regardless of seasonal growing season, yields greater experimental precision. For bean-to-pod lower coefficients of variation are in plots with 12 basic units (24 plants), for cultures performed in the greenhouse and field and 14 basic units for crops in the tunnel, the seasonal fall-winter season. In seasonal spring-summer season the plot size was 15 basic units for crops and the tunnel to the field. Seasonal stations in both the best grouping to use the harvest was complete. Theoretical models of spherical semivariogram, exponential and Gaussian estimates show weak spatial dependence. The average fresh weight of the fruit has random distribution within the farming environments not influenced plot size or number of harvests.