RESUMO
Six plots (three as controls) were established in southern Brazil to assess the reduction in runoff by vertical mulching (vm). In the other three, ditches (0.08-m wide, 0.4-m deep) were cut perpendicular to the slope, 2-m and 12-m upslope from the bottom of the plot where a flume was installed to measure runoff. Three simulated rainfalls (80 mm h-1 for 120 min, 80 mm h-1 for 65 min, 97 mm h-1 for 65 min) were applied with a sprinkler system to generate runoff. Compared to the controls, vm delayed its beginning by 10 to 20 min and reduced its volume by 34 to 39%, because initially all and later some runoff was caught in the ditches from where it infiltrated into the surrounding soil. Based on the experimental results, we derived a method to compute the runoff reduction by vm as influenced by ditch spacing (ds), initial soil moisture content (θi), and rainfall rate and duration. Computations for a combination of four ds (5 to 20 m), five θi (16 to 32%-vol), and three rainfall regimes (124 mm h-1 for 20 min, 66 mm h-1 for 60 min, and 42 mm h-1 for 120 min) showed an increasing runoff reduction as ds or θi decreased. In some cases, runoff was even completely prevented. The reductions were larger in the 60-min than in the 20-min rainfall. For the 120-min rainfall, there was no runoff with any combination of ds and θi.
