Soil compaction has led to reduced soybean yields, but the critical bulk density above which the yield decreases are poorly documented. Thus, this work aimed to evaluate the soybean yield (cv. Embrapa 48) in relation to the relative bulk density (Dsr) of a sandy loam (LVd) and a clayey Oxisol (LVef). A greenhouse experiment was carried out in 2001/02 with samples collected at the 0.0-0.2 m layer, sifted through a 4.0 mm sieve, fertilized and compacted in 0.03 m layers in 0.20 high and 0.25 m diameter pots. Four compaction and two water content levels (tension of 0.05 and 0.01 MPa) were evaluated in three replications. Soybean was sown in November (two plants per pot) and irrigated twice a day. In the second experiment on the field (2002/2003), only LVd was used and compacted by driving a 11 Mg tractor with four equally wide (0.40 m) tires, establishing five compaction levels, with four replications. Soybean was sown in December 2002, spaced 0.45 m between lines and 20 plants per meter. The bulk density (Ds) and maximum Ds was determined by the Proctor test, and the Dsr was obtained by dividing Ds by the maximum Ds. The soybean yield was evaluated at harvest. Under a soil water tension of 0.01, the optimum Dsr for soybean yield under greenhouse conditions was greater in the clayey eutroferric Oxisol (0.84) compared to the clay loam caulinitic Oxisol (0.75). The optimum Dsr for field soybean yield was 0.80.
soil compaction; soil management; Glycine max
