Soil management practices impacts on sorghum ( Sorghum bicolor) productivity have rarely been evaluated at field-scale. Field-scale soil management practices effects on sorghum grain yield were evaluated in three no-till crop-pasture rotation systems during two years in Uruguay (Oxyaquic Argiudoll). Treatments were established in a randomized complete block design in strips traversing the landscape in a sorghum-soyabean ( Glycine max) sequence integrated in three rotation systems: (1) continuous cropping (CC) with a winter cover crop of Lolium multiflorum; (2) short rotation (SR): two years pasture of T. pratense and L. multiflorum and two years of CC and; (3) long rotation (LR) four years pasture of Dactylis glomerata, Trifolium repens and Lotus corniculatus and two years of CC. Strips treatments included a factorial arrangement of two levels of cover crop residues (generated by winter grazing) with and without paraplough subsoiling. Strips were harvested with a combine equipped with a yield monitor. Data were analysed with mixed models accounting for spatial correlation. Yield was affected by year and rotation system but was not affected by management practices; either residue or subsoiling. Although its lower soil quality, CC had greater yield than SR and LR in 2006 (8.61 vs. 8.1 and 7.75 tonnes ha -1, respectively); however, no differences existed in 2007 (4.58 tonnes ha -1). Yield variations between field topographic zones were only found in 2007 SR (35%). Weak evidence of spatial correlation was found for soil properties at the site. No correlations were found between soil chemical properties and yield. Accounting for spatial correlation of 2006-2007 yields improved the statistical analysis. Animal treading and grazing did not appear to affect yield. For undegraded soils in temperate climates, cropping systems including no-tillage and perennial pastures preserved soil C, but did not guaranteed the same levels of grain productivity than more intensive cropping systems.
