Management effects on topsoil carbon and nitrogen in Swedish long-term field experiments - budget calculations with and without humus pool dynamics

摘要

The annual input and output of nitrogen (N) to and from agricultural land is relatively small compared with the total stock of organic N in soil. However, the long-term humus N pool dynamics is seldom considered in fertilisation planning, and thus N surpluses in budgets are considered as losses. In this paper, we analyse the consequences of this assumption and investigate how observed data precision affect the precision of model projections. We used available data sets from long-term agricultural field experiments in Sweden (topsoil C and N concentrations, crop yields as well as soil type and climate data) for calculating topsoil C and N mass dynamics. ICBM/2N, a simple C and N soil model (available at: www.mv.slu.se/vaxtnaring/olle/ICBM.html), was used for calculating soil organic matter balances. We parameterised this model for two field trials, and for specific crops, using available data and educated guesses and compared the results with those obtained from the conventional approach, not including humus pool C and N dynamics. In spite of the corrections for bulk densities, etc., the soil carbon measurements were too variable for a critical model validation or a sensitive test of the effects of including pool dynamics. In other words, we had to rely on the model assumptions for the projections and soil data could only be used to obtain general means, e.g., for the whole duration of the experiment. We think this is a general problem, not limited to this data set and model. We also show, at least in principle, how estimates of organic N pool dynamics can be used to produce improved N balance sheets for individual crops. For example, the apparent N use efficiency by sugar beet increased from 58 to 99% when organic N pool dynamics were included.