Effect of mechanical weeding on soil erosion and earthworm abundance in sugar beet (Beta vulgaris L.)

Autor/in: Fishkis, O., H.-J. Koch
Jahr: 2023
Zeitschrift: Soil and Tillage Research 225
Seite/n: doi.org/10.1016/j.still.2022.105548
Stichworte: Mechanical weed control Hoeing Soil erosion by water Earthworm abundance

Drucken

Abstract

The use of herbicides for weed control in plant production is widely criticized because of their potential aquatic and terrestrial toxicity. However, the environmental risks of mechanical weed control as an alternative control method in row crops have not yet been studied in Europe. Field trials in Southern Lower Saxony were carried out over three years on five sugar beet fields to assess the effect of weeding by hoeing on earthworm abundance, runoff, and soil erosion. Treatments were hoeing (H), herbicide spraying (S) and combined hoeing-spraying (HS), tested in four replicates at each field. The abundance of adult and juvenile earthworms of anecic and endogeic groups was determined in autumn using formalin expulsion 4 months after the final weed control operation. Runoff and soil erosion in mechanically and chemically weeded plots were measured by means of rainfall simulation on 2-m2 subplots after weed control. The experiment was organized in completely randomized block design. The abundance of adult and juvenile earthworms of anecic and endogeic groups was not affected by the weed control treatment. Cumulative runoff and soil loss were significantly reduced by mechanical weed control by a factor of six and eight, respectively, but only if the soil was crusted before weed control. In the absence of a soil crust before weed control there was no difference in runoff and erosion between mechanical and chemical weed control. Measured runoff and soil loss were used to calculate the curve number (CN) and the soil loss ratio (SLR) which are the empirical parameters for the prediction of runoff and soil erosion with the Curve Number SCS method and the Revised Universal Soil Loss Equation, respectively. CN was significantly increased and SLR tended (p =0.08) to be increased if the soil was crusted before weed control. Mechanical weed control reduced CN and SLR in crusted soils but had no effect in non-crusted soils. Therefore, for correct runoff and soil erosion prediction the effect of crusting and mechanical weeding on CN and SLR must be considered. In conclusion, mechanical weeding was revealed to be particularly beneficial for crusted soil by simultaneously reducing toxicological and erosional risks.

FaLang translation system by Faboba