Soil chemical properties affecting grain yield and oil content of crambe biofuel crop

Crambe (Crambe abyssinica Hochst) is an important grain crop for biodiesel production. The objective of this study was to evaluate how spatial distribution of soil properties affects crambe grain yield and oil content. We analyzed six different sampling densities, with a distance between sampling points ranging from 5.5 to 44 m, to identify correlations among soil and crop variables, and the best spatial distribution. Crambe oil content was not correlated to any of the physicochemical properties, whereas crambe grain yield was correlated with total porosity and exchangeable Al and Ca, effective cation exchange capacity, available P, K, Cu, and Zn, and aluminum and base saturation, with highly significant correlation (p < 0.001) for P and K and Cu and Zn. Most soil properties had spatial dependence degrees from weak to strong, and were modeled by spherical or exponential semi-variograms. Spatial data showed that low grain yield is associated with high seed oil content, while crambe grain yield was more correlated with phosphorous availability in soil in the first year (2012) and with Ca, Mg, and soil organic matter in the second year (2013). Spatial variability in soil nutrients is attributed to uneven distribution of fertilizers and surface runoff. The results demonstrate sampling density must be increased to generate more accurate maps for soil chemical properties affecting crambe. The use of a grid of 25 points in a 625 m(2) of land is suggested, providing more than 60% accuracy for grain yield estimation and more than 40% for most chemical properties' estimation.