Models of soil water erosion prediction and estimation are generally based on a series of parameters, normally determined by applying simulated rainfall in the field or laboratory. The use of simpler, faster and cheaper parameters such as texture, content and oxide type and aggregate water stability may contribute to facilitate the estimation of the factor interrill soil erodibility (Ki). The purpose of this study was: i) to determine the interrill soil erodibility of three Brazilian Oxisols; ii) evaluate the use of soil sand and clay contents in the prediction of interrill soil erodibility, as in the WEPP model; iii) verify the correlation among some soil properties and the interrill soil erodibility factor. The interrill soil erosion rates were determined in laboratory, using three Oxisols with different clay content of the State of Rio Grande do Sul, Brazil, under a mean intensity of simulated rainfall of 88 mm h-1, applied during 180 min on bare soil, using experimental interrill plots of 0.36 m² at a slope of 0.09 m m-1 and an initial water tension of 6 kPa. The susceptibility of soils with high clay and Fe oxide contents and higher water aggregate stability to interrill erosion was lower. The experimentally determined values of interrill soil erodibility (Ki factor) varied from 0.76 x 10(6) kg s m-4 to 1.48 x10(6) kg s m-4 for the studied Oxisols. The aggregate water stability index was highly correlated with the interrill soil erodibility factor (r= -0.90**). The use of soil sand and clay contents, as suggested by the WEPP model, did not prove adequate to estimate the interrill soil erodibility (Ki factor). The variation in clay and Fe oxide contents explained 97 % (p<0.01) of the interrill soil erodibility variation.
water erosion prediction; soil parameters; aggregate stability
