Spatial variability of soil properties and their association with the landscape position are needed in the application of site specific-management practices. Crop yields are highly variable across a field as a result of complex interactions among different factors such as topography, soil attributes and management practices. The objetives of this study were to determine the spatial distribution of the available water storage capacity using pedotransfer functions and to identify soil attributes associated to the variability of corn yields in a field scale in the Sandy Pampas. In two corn fields, 8 ha and 10 ha-plots were marked and yield maps were obtained at harvest. Before corn planting, a geo-referenced sampling following a grid design was carried out. Thirty-two cores were collected from 0-30 cm-depths in Site 1 (8 ha) and forty-two cores in Site 2 (10 ha), and CO and texture were determined. Available water storage capacity (CAD) was estimated using pedotransfer functions. In three representatives soil profiles (Typic Hapludoll, Entic Hapludoll and Entic Hapludoll, convex phase) water retention at -33 kPa and -1,500 kPa, texture and CO was determinated in order to determine the pedotransfer function with the best fit. The results of both sampling grids were analyzed through geostatistical procedures. CAD in the upper meter of the soil profiles was 121 mm in the Typic Hapludoll and 78-79 mm in both Entic Hapludolls. Sand content variability between topographic positions was greater in Site 1 (40-81 percent) than in Site 2 (43-73 percent) showing a moderate spatial structure. CAD had a moderate spatial structure in Site 1 but not in Site 2. Corn yields showed a strong spatial dependence in Site 1 but only a moderate spatial dependence in Site 2, fluctuating between 5,4-14,5 t ha-1 and 5,5-13,3 t ha-1 in both sites, respectively. Our results show that the sand spatial variability is closely related to corn yields, explaining 64 percent of corn yield (Y) variability: Y (t ha-1) = 21,5 - 0,189 Sand (percent) (P less than 0,001) n=74. In spite of the high association between CAD and corn yields, the high cost of determination of the former and the data required to use pedotransfer functions could limit the possibilities of its use to determine the potential production of corn. In a dry year (Site 1) with similar CAD levels, corn yields were 37 percent higher in foot and mid-slopes dominated by Entic Hapludolls compared to those at the hilltops. These results could be attributed to the water table rise in the former soils.
Urricariet, A. S., Niborski, M. J., & Kvolek, C. M. (2011). Atributos del suelo y paisaje asociados a la variabilidad de rendimientos de maíz en la Pampa arenosa. Ciencia del Suelo : Revista de la Asociación Argentina de la Ciencia del Suelo,29, (1),p.49-59
Urricariet, Alicia Susana, Niborski, Mauricio José, Kvolek, Claudio Miguel. 2011. "Atributos del suelo y paisaje asociados a la variabilidad de rendimientos de maíz en la Pampa arenosa". Ciencia del Suelo : Revista de la Asociación Argentina de la Ciencia del Suelo 29, no.1:49-59.
Recuperado de http://ri.agro.uba.ar/greenstone3/library/collection/arti/document/2011Urricariet