The surplus of the NPK balances of Hungarian agriculture, appearing in the nutrient balance sheet is a consequence of intensive fertilization in the past. Hungary had a positive P-balance between 1960 and 1990 for 30 years, and a positive NK-balance between 1970 and 1990 for 20 years during the 20th century.Studying the long-term effects of superphosphate in a field experiment on calcareous chernozem soil, it was found that the “half life time” of residual P was 5–7 years. The trial, i.e. the P after-effect lasted for 20 years.Liming and fertilization are the main soil fertility improving considerations for acidic sandy soils. These soils are often poor in all major nutrients. Mineral fertilizers, mainly NH4NO3, acidify soil, the pH values sink in comparison to the unfertilized plot. Liming and use of dolomite powder may counterbalance the acidification.
Increasing doses of farmyard manure (FYM) or equivalent mineral NPK fertilizers and their combinations were analysed in a crop rotation with potato, maize and winter wheat with special regard to their long-term influence on soil fertility. The yield-increasing capacity of FYM doses was only 82%, as compared to the equivalent amount of mineral NPK. Fairly high N release (50.9 kg ha
) could be observed on the unfertilized plots. Great differences in N utilization were recorded, depending on the form and dose of fertilizers. The average N utilization from FYM was only 29.3%, while that of the equivalent fertilizer application was 49.8%. The lowest soil reactions were observed without fertilization and at the highest NPK doses. Negative N balances generally resulted in low soil organic matter content. FYM and equivalent NPK fertilizers had a similar influence on the ammonium lactate (AL)-extractable K
O content of the soil, while an increase in the AL-P
content could be observed in the case of mineral fertilization.
Old-fields (44, aged 1–15 years, from Czech Republic and Hungary) were sorted according to their soil moisture and nitrogen content into wet, mesic or dry, and nutrient poor, moderate or nutrient rich categories, resulting in 8 combinations (dry and nutrient rich fields were not present). The vegetation of old fields was sampled using phytosociological relevès. The changes in species cover data and importance of species trait categories were analysed in relation to three environmental factors, i.e., time since abandonment, soil moisture and total soil nitrogen using ordination, generalized linear models (GLM) and regression tree methods. Successional seres in the first 15 years after field abandonment were divergent. Species diversity significantly decreased with increasing site moisture and was highest in sites with moderate nitrogen content; while the relationship with time was not significant. Raunkiaer life forms and life strategies (sensu Grime) were generally the most predictive species traits considering species occurrence during the course of succession, the type of dispersal considering the different moisture status, and the ability to lateral spread considering the nutrient status of the old-fields. Most trends appeared in both parametric GLM and non-parametric regression tree analyses, several only in GLM. We consider regression trees to be a more convenient tool than GLM in cases such as ours with a rather small number of samples and robust character of data. Another advantage is that a hierarchy of species traits is taken into account. Thus, the occurrence of a species along an environmental gradient can be predicted if the species possesses a certain combination of traits.