Results of soil survey, water and tillage erosion studies carried out on two slopes in small loess agricultural catchment of the Lublin Uplands are presented in the paper. Detailed soil survey carried out on two slopes of the catchment enabled to localize soils of various classes of erosion. Using soil profile reconstruction method, total erosion and accumulation volume was calculated for slopes during the whole period of their agricultural use. Annual erosion rate was 3.94 mm for 7% slope (at 70 years of agricultural use) and 2.87 mm for 12% slope (at 170-190 years), whilst erosion diminished by accumulation, 2.46 and 1.66 mm, respectively.

Studies of water erosion contained measurements of splash, and runoff and soil loss from a system of plots of variuos length (2.5, 5, 10 and 20 m), and tillage erosion – measurements of soil translocation due to shallow plow (10 cm) and cultivator (5 cm). Splash measurements were carried out with splash cup of various diameters (3.2, 5.0, 7.3, 11.7 and 14.5 cm). To calibrate splash amounts, the exponential theory of splash distribution of Van Dijk et al. (2002) was applied.

Results of splash studies showed that splash amount depended on rainfall parameters, and splash distance on micro-relief of the soil surface and presence of running water. To evaluate splash amounts, a simple model was developed on the basis of relation between splash and 10-minute rainfall intensity and tested. Results of measurements showed that splash amounts were many times larger in comparison to measured soil loss. Only in the case of rill erosion, splash amounts were lower than soil loss.

To describe the process of water erosion, a new interpretation method was proposed to analyze soil loss from plots of various length. The interpretation is based on maximum soil loss per unit area, as the value characteristic of the erosion event, and the final ratio of soil loss from plots, as the reflection of time during which eroded soil material can reach the plot outlet. The time was well correlated with the effective distance of soil transport. Analysis of results showed that time to reach plot outlet was on average 1.8-times longer for 7% slope in comparison to 12% slope. It implies that soil loss can reach the main water course of the catchment in different time, not only being dependent on the distance but also on slope. Water erosion rates were best related to rainfall and runoff erosivity index EI30 of the USLE. However, the relation was of general character. The studies confirmed that the USLE principles are valid only inside the effective distance for soil transport, which is most probably connected with time of rainfall and runoff.

Studies of soil displacement due to tillage implements enable to compare soil loss due to tillage and water erosion during plant vegetation. Two various geomorphic aspects were compared - lowering of slope and restructuring of lower slope border. Comparison showed that tillage erosion rates were much higher than water erosion rates during vegetation. Both studied processes were 15% and 62% of total annual erosion rates that were assessed from soil profile reconstruction. Low percentage of erosion share of studied processes on slope of southern exposure (15%) was most probably connected with erosion from snowmelt (not measured in the studies). Tillage erosion was found to be more important for restructuring of lower slope border, and only extreme events during sugar beet vegetation or leaving the field without canopy cover after cereals harvest can be close to the effect of tillage.