Conventional tillage systems with high soil disturbance are being steadily replaced by tillage systems with low or no soil disturbance. An approach using three methodological steps (greenhouse, deliberate seed burial and field) revealed the long-term vertical distribution and losses of a soil seed bank as effects of different tillage operations. Seeds (oilseed rape; Brassica napus L.) and seed substitutes (plastic pellets) acted as models for a seed bank. (a) A pot experiment in the greenhouse showed that emergence rates were highest in soil depths of 1-5 cm. Germination and emergence was clearly reduced in depths of 0 and 7 cm, and emergence was completely inhibited at 12 cm. About 40-50% of seeds fell dormant in 0 and 12 cm depth, while almost no seeds fell dormant in 1-7 cm depth. (b) The high-dormancy variety Smart persisted to a high extent (60% of the initial seed number), but only 8% of seeds of the low-dormancy variety Express persisted over 4.5 years, after deliberate seed burial. Seed persistence was similar in all soil depths of 0-10 cm, 10-20 cm, and 20-30 cm. (c) The field experiment lasted from 2004 to 2009 and had different tillage treatments of inversion and non-inversion tillage: stubble tillage immediately after harvest combined with primary tillage by mouldboard plough (SP), chisel plough (SC), or rototiller (SRTT); primary tillage without stubble tillage by mouldboard plough (P), chisel plough (C); or no tillage (NT). The seed bank from an artificial seed rain of 20,000 seeds m -2 was significantly higher in all treatments with immediate stubble tillage, and clearly declined over time. However, seed bank depletion was slow once a seed bank had been established. The distribution of oilseed rape seeds and plastic pellets (7000 pellets m -2 broadcast) tended to equalise over the soil layers of 0-10, 10-20 and 20-30 cm over the course of five years. Since seed bank depletion was not attributable to a specific soil depth, shallow and low disturbance tillage did not generally result in a high seed persistence. More important than the depth was the timing of tillage. Though no-till systems provided conditions for seeds to fall dormant at the soil surface to a small extent, the effect lasted only for a limited time. Seed substitutes can be well used in methodological approaches to picture movement of seeds in the soil in order to optimize tillage strategies in agricultural practice.
