Sie sind hier: Startseite Bonner Bodenk. Abh. Band 31 (2000) - Band 40 (2005) Band 39 (2003) Summary Band 39

Summary Band 39

Schmelmer, Karin: Soil erosion processes, surface runoff reduction and sediment trapping effectiveness of grass buffer strips – Field measurements and application of prediction models. Bonner Bodenkundl. Abh. 39 (2003), 267 p.


In the years 1997 - 1999 field studies were conducted to determine the surface runoff reduction and sediment trapping effectiveness of grass buffer strips. The 10 % inclined field was conven-tionally tilled with maize. Measurements were realized making use of natural and simulated (40 mm/h) rainfall. To evaluate the time variable erodibility of the soil and to characterize buffer strip input, examinations of aggregate stability and splash-erosion were added.

Aggregate stability was at maximum in summer with moderate soil moisture. A soil moisture of field capacity or higher caused a heavy decrease of aggregate stability. The lowest values were obtained in winter. Aggregate slaking due to entrapped air compression caused moderate to regu-lar soil surface sealing in summer. In winter regular surface sealing was the result of silt disper-sion. The most stable aggregate breakdown products were those of 200 - 630 µm diameter. De-creasing aggregate stability was characterized by a decline of aggregate fragments > 2000 µm and an increasing portion of fragments 63 - 200 µm and those < 63 µm.

Due to the lower kinetic energy, simulated rain detached lesser and finer soil material than natu-ral rain did. The natural rain parameter NI2 determines the amount of detached soil. The different particle sizes show an individual relationship to rain intensity, depending on their stability and their portion of primary and aggregated particles. With aid of erodibility coefficients, which were calculated for different soil conditions, total amount and particle size distribution of splash-material are predictible. Kinetic rain energy is less appropriate for the prediction of total splash amount than rain intensity is. The prediction of particle size distribution is not possible using kinetic rain energy values.

To cause surface runoff on antecedent unsealed soil, 10 min-intensities of ≥ 50 mm/h and 30 min-intensities of ≥ 30 mm/h were necessary. High intense rain ≤ 10 mm total amount caused surface runoff if the peak intensities occurred late in the rainfall event, when infiltration capacity was de-creasing. Because of its lower compressive force, a higher amount of simulated rain was needed for surface sealing and runoff to cause as compared to natural rain. The higher erosivity of natural rain and a greater plot length resulted in higher runoff rates and sediment concentration as well as a coarser sediment load as compared to the rainulator-runoff events.

Grass buffer strips reduced surface runoff by 45 % if soil moisture was about field capacity, and up to 100 % at a antecedent soil moisture up to 30 Vol.-%. Doubling the grass buffer length furthered surface runoff reduction by additional 15 %. During small runoff events a “field border strip” (an annual plant cover consisting of grass and herbes, which could grow on the field area due to renunciation of herbicides) with a plant cover of 60 % offered a comparable effectiveness like a permanent grass buffer strip with a plant cover of 90 - 100 %. During one greater runoff event the surface runoff reduction by “field border strips” with a plant cover of 40 % and 60 % amounted only 10 % and 44 %, respectively. Sediment removal by “field border strips” was sig-nificantly higher than surface runoff reduction. During simulated rain grass buffer strips caused a decrease of sediment concentration by 85 %, independend on the concentration of the incoming sediment. Animal burrows caused highly variable surface runoff retention, while sediment remo-val was hardly affected. Also the buffer strip length did not influence sediment removal. Total sediment load was removed mostly by 88 - 100 %. The load of fine silt and clay was reduced by 80 - 100 %. In a flow distance ≥ 1 m the sediment deposited in the grass buffer area contained no more primary sand. The relative enrichment of finer primary particles with increasing flow dis-tance was the effect of an enrichment of middle sand sized aggregates. In a flow distance ≥ 3 m the sediment trapping effectiveness of grass buffer strips decreased largely.

The higher incoming surface runoff rates of a natural runoff event caused a more concentrated flow inside the buffer strip and therefore a lesser portion of the buffer area to be involved in redu-cing runoff amount by infiltration as compared to the rainulator-runoff events. Consequently the surface runoff retention performance of the grass buffers appears less during natural events, pre-suming similar antecedent soil conditions for the natural and rainulator-runoff events.

The event based soil erosion model EUROSEM simulates the surface runoff retention by grass buffer strips very good for high antecedent soil moisture conditions or for concentrated buffer strip effluent. It overestimates the infiltration decline for dry antecedent soil conditions, so that surface runoff retention is underestimated for these cases. Sediment removal was somewhat over-estimated for the rainulator-runoff events and underestimated for the natural one. Due to the lo-wer soil compaction by simulated rain the artificial and the natural runoff events cannot be simu-lated with OPUS/OPUS_2 in the same simulation run. OPUS_2 is not yet officially released. It simulates surface runoff reduction by grass buffer strips properly. Sediment trapping however is overestimated for the natural runoff event. For a rainulator-runoff event the model calculates se-diment concentration reduction corresponding to the measured values.