Browsing by Subject "N-Mineralisation"

Now showing 1 - 1 of 1

Einfluss der Bearbeitungsintensität beim Umbruch von Luzerne-Kleegras auf die Stickstoffmineralisation zur Folgefrucht Winterweizen im organischen Landbau
(2003) Wald, Fabian; Claupein, Wilhelm
In the crop rotation of organic farming grass-legume mixtures play an important role due to the legumes´ ability to assimilate N. Ploughing-in of established grass-legume mixtures results in releasing the assimilated N by mineralisation of organic matter. In practice the mineralisation can only be controlled by means of soil cultivation. The aim of the present study is to analyse the relations between different intensities of soil cultivation and N-mineralisation. The data were used to test the simulation model CANDY. The field experiment of each 0.1 ha was set up at three sites in two different locations, which were cultivated from 1999 to 2001: Hohenheim (with trials 610 and 611) and Kleinhohenheim (with trial 660). In the beginning all sites had a three-year old grass-clover-alfalfa mixture, which was ploughed-in for trial 610 and 660 in the late summer of 1999 and in the year 2000 for trial 611. The factor soil tillage was varied in three stages as follows: RT+RT+plough: double rototill cultivation (RT, 10 cm deep) in intervals of approx. 2 weeks, followed by ploughing (plough, 25 cm deep); RT+plough: single rototill cultivation, followed by ploughing (depths as mentioned above); Plough: ploughing without any preceding cultivation (depth 25 cm). After uniform seedbed preparation with a rotary harrow, wheat was sown on all trial sites in autumn, and in trials 610 and 660 it was followed by oat in 2001. Nitrogen content in the soil was determined by repeated sampling at a depths of 0-10, 10-20, 20-30, 30-60 and 60-90 cm. Monitoring boxes were installed in 1 m depth in an undisturbed soil body from September 2000 until April 2001 to record nitrate leaching. Ploughing-in of the grass-clover-alfalfa by means of rototiller cultivation (treatments RT+RT+plough and RT+plough) was followed by a significant increase of mineralisation, which in case of the plough treatment was less pronounced. In this case the date of cultivation, 6 weeks after the rototilling, may have had an influence. Nmin-contents in autumn 1999 were higher after RT+RT+plough than after RT+plough. It has to be taken into consideration that there was a time gap between both treatments of 9 days. But also in the following year (611), when both treatments were cultivated the same date, there was a significant, slight difference of the Nmin values depending on the treatment. Nitrate leaching was only measured in trial 611. Quantities of 86, 84 and 64 kg N/ha were observed in treatments RT+RT+plough, RT+plough and plough, respectively during winter. Due to high Nmin-contents in autumn, for the rototill treatments higher nitrate losses can be assumed compared to the plough treatment for both years of experiment. Depending on the location, nitrogen uptake and yields of wheat turned out to be different. In Kleinhohenheim they were lower in treatments RT+RT+plough and RT+plough than in the plough treatment. It was the other way round in Hohenheim on a higher production level. Due to strong hail impact, this relation between the treatments was not to be proved in trial 611. Oat was the second crop. In this case no effects of intensity of soil cultivation on nitrogen uptake and yield could be observed between treatments and locations. The CANDY model was used for simulating the results of trials 610 and 611. First, the model seemed to be inadequate because it could not model the N-dynamic after soil cultivation. Adding fictitious organic material to the system helped to overcome this problem and then, on average, the N-dynamic model fit was satisfying. An estimate to overcome the general insufficient fit of the model could be mineralisation of parts of the physically protected organic matter (SOS), which is already implemented in the model, right at the moment of cultivation. Data of soil moisture of trial 611 served to calibrate the model successfully. With amended soil parameters the model was then easily applied to the corresponding data of trial 610. In contrast, CANDY did not predict well the nitrate leaching - possibly because the model did not consider preferential flow.