Tillage caused dispersion of phosphorus and soil in four 16-year old field experiments

Abstract

Long-term field experiments are important for studies of the long-term effects of agricultural management practices. Unfortunately tillage caused dispersion of soil from plot to plot is a serious problem in such experiments if the plots are unconfined and tillage takes place across plot borders. The extent of this problem is only documented in few relatively old field experiments and these do not reflect present day tillage operations. In this study, four 16-year old field experiments with different phosphorus fertilisation treatments were used to quantify the contemporary extent of this problem. A two-dimensional dispersion model fitted well to measurements of total soil P content in transects across and along plots of the four experiments. We found that tillage caused soil dispersion across and along the plots on average were 0.34 and 0.72 m 2 per tillage year. This is significantly higher than found in previous studies, reflecting that contemporary tillage operations move soil more around than previous tillage practices. Already after 5 years of tillage, 3–11% of the net added P had left the plots. After 15 or 16 years of experimentation, at the time of soil sampling this had increased to 14–18%. We therefore conclude, that contemporary tillage operations move soil around to an extent, which is not compatible with experimental designs having no permanent borders between plots and we recommend that designs of new long-term field experiments take this into account. Regarding existing old long-term field experiment with unconfined plots, it is important to acknowledge the fact that soil movement between plots has taken place and that accumulated treatment effects therefore are seriously blurred. Relating treatments to responses in soil and crops may therefore be seriously wrong. However, it is still possible to utilise the actual differences of measured parameters between plots in such experiments and relate these to each other. Therefore, in spite of the problems with soil dispersion between plots, such old long-term field experiments still play an important role as living soil archives providing important material and information for a wide range of process-oriented studies.