Effects of stubble management, tillage and cropping sequence on sustainable wheat production in the South-Eastern highlands of Ethiopia

[摘要] English: Bread wheat (Triticum aestivum) is one of the major cereals produced in south-eastern Ethiopia. Yields are often low on peasant farmers' fields due to sub-optimal crop management practices. Four multi-factor crop management trials were conducted which ran from 1992 till 2000 at Kulumsa and Asasa in the south-eastern highlands of Ethiopia. Two trials were conducted at each location, one where an ox-plow was used to simulate the peasant farmers methods and on the other trial mechanized farming methods were used. Different crop residue management options, tillage practices and cropping sequences were included as treatments. The crop residue management treatments were burning, partial removal (to simulate grazing by animals) and complete crop residue retention. The tillage practices for the mechanized trial were conventional mouldboard plowing, conservation tillage which was zero tillage at Kulumsa and mimimum tillage at Asasa and for the ox-plow trials it were conventional ox-plowing and minimum tillage. The cropping sequences were continuous wheat and a rotational system of faba beans (Vicia faba) followed by two seasons of wheat. The soil types were a clay intergrade between an eutric Nitisol and a luvic Phaoezem with 50% clay in the topsoil at Kulumsa, and a clay loam calcic Chernozem with 36% clay in the topsoil at Asasa. The objective of the study was therefore to determine the integrated effects of cropping sequence, straw management and tillage practices on the productivity and sustainability of wheat-based farming systems in Ethiopia. Among the crop residue management treatments, burning of stubble tended to increase the grain and biomass yield, as well as the yield components of wheat in most of the years when compared to partial removal and complete retention of stubble. Tillage did not give consistent responses for most of the parameters in either of the trials, with the exception of thousand kernel weight which was consistently higher for zero or minimum tillage in most seasons. Conventional tillage tended to be the superior tillage practice in terms of most of the measured parameters. The faba bean cropping sequence markedly increased wheat yields and the yield components, especially in the first wheat crop following faba beans. This can be ascribed to a higher soil N status. Burning of stubble tended to also enhance wheat N uptake as compared to partial removal or complete retention of the straw. Stubble management and cropping sequence had relatively minimal effects on soil strength expressed as penetrometer resistance. Conservation tillage led to a higher penetrometer resistance particularly in the surface layer of the soil (i.e., 0-15 cm depth), in both the mechanised and ox-plow systems, compared to conventional tillage. The penetrometer resistance of the 0-5 cm depth was negatively related to grain yield, while the penetrometer resistance of at the 20-25 cm depth was positively related to grain y.ield. Retention of straw on the soil surface increased the concentration of most of the important plant nutrients in the upper layer (i.e., 0-15 cm) of the soil as compared to burning and partial removal of the straw. Zero or minimum tillage practices also increased the concentration of some important plant nutrients like phosphorus and potassium in the upper layer of the soil in both system trials. Cropping sequence had little effect on the soil chemical properties. Partial removal or retention of the stubble tended to increase the population density of some broadleaf weed species, while burning had the opposite effect. Burning of crop stubble also markedly reduced the total grass weed population as compared to the other straw management treatments. Broadleaf weed populations were not affected by tillage practices in either the ox-plow or mechanised trials. Grass weeds, however, increased significantly in density under minimum or zero tillage. Broadleaf weeds did not vary markedly in response to cropping sequence, but most of the grass weed populations decreased in the faba bean rotation. A three years rotation, consisting of two consecutive wheat crops followed by one faba bean crop reduced the incidence of take-all, but had little effect on the eyespot incidence. Both zero or minimum tillage dramatically inhibited the take-all incidence relative to conventional tillage, but had little effect on eyespot incidences. Stubble burning had no consistent effect on the incidence of either of the diseases. The incidence of eyespot was higher in the full stubble retention practices compared to partial removal. A partial budget analysis reflected that the practice consisting of faba beans in rotation with wheat, combined with stubble burning, was the lowest cost option for ox-plow production systems. The practice of a faba bean-wheat rotation plus full stubble retention was the lowest cost option for mechanised production systems. The economic optimum practice' for both the mechanised and ox-plow systems at Kulumsa was combining stubble burning with conventional tillage and continuous wheat. However, for the Asasa mechanised trial the economic optimum practice was combining stubble burning with conventional tillage and a faba bean-wheat crop rotation, while for oxplowing it was combining stubble burning with minimum tillage in a faba bean-wheat crop rotation. Moreover, the index of variability of the nett benefit of each trial did not vary markedly between the low-cost and the economic optimum treatments. This means that the stability of farm level income will not be affected significantly by adopting the economic optimum practices for each zone. Pot experiments were conducted to determine the effect of the different straw management practices on wheat seedling development under controlled conditions. Straw retention had a negative effect upon most of the measured seedling growth parameters. The nil straw treatment (i.e., the control) grew better than the other application levels. Straw retention levels higher than 2.5 t ha-! impeded seedling development. Burning of the straw reduced the negative effect of straw application. The incorporation of the straw in to the soil resulted in the most pronounced inhibition of seedling development. The application of fresh straw was more harmful than naturally degraded straw.

[发布机构] University of the Free State