Tillage and cropping effects on selected properties of an Argiudoll in Argentina

Abstract

Soils of the Argentine humid pampa region are usually weakly structured due to its high silt content. Selecting crop sequence or tillage systems are an alternative in small farms for the protection of the soil against physical degradation and erosion given that conservation practices, grass meadows, and fertilizers are expensive and therefore rarely used. Evaluation of selected soil properties was conducted on soil sampled from a long‐term tillage experiment with continuous soybean established in 1975 on a Typic Argiudoll silty loam soil in Argentina. Tillage treatments included conventional tillage with moldboard plow (CT), chisel plow (CP), and no till (NT). A comparison with continuous corn under NT was also carried out. Sampling was performed after the emergence of both crops in 1990. Tillage and cropping treatments affected properties related to soil slacking and dispersion to a greater extent than they did on aggregate size distribution. According to the De Leenheer and De Boodt index, aggregate stability within soybean soil classified as bad for CT, unsatisfactory for CP, and very good for NT, whereas the soil with corn under NT classified as excellent. The no tillage treatment within soybean had significantly more organic carbon in the 0–5 cm depth than CP or CT. Soil respiration was significantly higher in NT than in CT in the surface layer, while CT showed higher values in the 10–15 cm depth. Tillage treatments did not significantly affect microbial biomass under soybean cropping. The effect of monoculture corn versus monoculture soybean under NT on soil respiration, biomass and organic carbon was not significant. Soil pH in the 0–5 cm depth under soybean was in the order NT > CP > CT, whereas the soil with corn under NT was more acid than the soybean soil (P=0.05). Cation exchange capacity and exchangeable bases followed a similar trend. Organic carbon (0–5 cm depth) and aggregate stability were significantly correlated when samples from all treatments were considered.