The impact of different NPK supply levels and their combinations on the yield, mineral nutrient content and element uptake of an eight-component, legume-free grass mix with meadow fescue (Festuca pratensis) as the main species was examined in 2010, the 37th year of the experiment and the 10th year of the grassland. The soil of the production site contained 3% humus, 3–5% CaCO3 and 20–22% clay. The N and K supply of the site was originally moderate, but the P and Zn supply was relatively low. The experiment consists of 4N × 4P × 4K = 64 treatments × 2 replications = 128 plots. Groundwater level is 13-15 m and the area is drought sensitive. In the examined year of 2010, there was 319 mm precipitation in the area until the first reaping. Despite the plentiful subsequent precipitation, the usual September reaping did not take place in the aging 10-year-old grassland. Main results:

The botanical composition of the population was mainly determiend by the N×P fertilisation. The coverage of crested wheatgrass was 37% and that of smooth brome was 27% on the soil which was well supplied with both nitrogen and phosphorus. Tall fescue and cerastium profilerated in treatments moderately supplied with N. The coverage of legume weeds (mainly black medic, vetch and coronilla vetch) increased over 16% in plots not fertilised with N. As a result of N fertilisation, grass coverage increased from 13% to 66%, while weed coverage decreased from 32% to 16% and the total coverage changed from 45% to 82%. At the same time, the average number of grass species increased and the number of weed species decreased.

The hay mass increased sixfold as a result of N fertilisation and nearly sevenfold as the joint result of NP fertilisation. The N, Mn and Cu content of hay increased with the N supply, its P, S and Sr content increased with the P supply and the K content increased wiht the K supply. The concentration of other examined elements became more diluted in the increasing yield. As a result of the N×P interactions, the following ranges were observed: P 0.12–0.31%; Na 25–1302 mg kg-1; Sr 9–20 mg kg-1; Mo 0.4–1.8 mg kg-1; N×K treatments: K 0.86–1.84%. Elements taken up with yield had the following rnages in the N×P treatments: N 8–132 kg, Ca 3–30 kg, P 2–13 kg, S 1–12 kg, Mg 1–8 kg, Na 0.1–4.2 kg per hectare. N×K treatments: K 8–103 kg, Zn 10–138 g, Al 48–243 g, B 7–20 g, Ni 0.4–1.8 g, Co 0.1–0.4 g per hectare.

Similarly to the ammonium lactate + acetic acid method, the NH4-acetate+EDTA soluble P content well reflected the fertilisation level and nutrient supply condition of the soil. The obtained data can be obtained in the following way: NH4-acetate+EDTA-P × 1.7 = AL-P. The two methods provide identical nominal soluble K content values. For this reason, both methods are equally suitable for the determination of the K condition of the soil in consulting. The data obtained with the NH4-acetate+EDTA method do not need specific calibration in terms of supply level limit values. The soluble Al and Sr concentration increased in the ploughed layer by applying superphosphate doses, whiel the N and Sr content decreased.

Keywords: established all-grassland, NPK fertilisation, botanical composition, element uptake