No-till is widely used to control soil erosion in agricultural areas in Brazil and is currently practiced on about 30. Mha. However, studies have shown that no-till is not as efficient in controlling surface runoff losses as it is in reducing soil loss. The objective of this study is to evaluate soil and surface runoff losses on small and large plots with differing slope lengths, cropping sequences and tillage systems in southern Brazil. Surface runoff and soil losses under natural rainfall erosion plots (3.5. ×. 11. m, 3.5. ×. 22. m, 50. ×. 100. m, and 100. ×. 100. m) were evaluated in two experiments in a well-drained Oxisol (>60% clay) with 9% and 7% slopes, respectively. The experiment extended over 14 years comparing 4 different soil management systems: (a) bare soil plots with slope length 22. m; (b) bare soil plots with slope length 11. m; (c) sequence of wheat (Triticum aestivum)/soybean [. Glycine max (L.) Merr] with disk plow. +. lighter off-set disk-harrow (DP+LD); and (d) sequence of wheat/soybean under no-till (NT). In another experiment using large field plots, three soil tillage regimens (DP. +. LD; heavy off-set disk-harrow. +. lighter off-set disk-harrow (HD. +. LD), and NT) were compared over the course of a 5-yr. crop sequence of black oats (Avena estrigosa)/soybean-black oats/corn (Zea mays L.)-wheat/soybean- black oats/soybean -blue lupine (Lupinus angustifolium)/corn. Results for both experiments show that, when compared with conventional soil tillage (DP. +. LD or HD. +. LD), soil losses for NT were >. 70% lower. However, the benefit of reduced surface runoff losses was less evident, suggesting the need to implement additional practices to control surface runoff to avoid transport of pollutants to waterways. © 2015 Elsevier B.V.
