The effect of various levels of NPK supplies and their combinations on the element uptake, oil content and fatty acid composition of winter rape (cv. Yet Neuf), and on the NO3-N and SO4-S leaching and salt balance in the 0-300 cm soil layer was examined in the 11(th) year of a long-term mineral fertilisation experiment set up on calcareous loamy chernozem soil. The ploughed layer of soil at the growing site contained approx. 3 la humus, 5% CaCO3 and 20% clay. Soil analyses showed that the soil was moderately well supplied with N, K and Cu and poorly supplied with P and Zn. The experiments involved 4N x 4 P x 4K = 64 treatments in 2 replications, giving a total of 128 plots. The mineral fertiliser was applied in the form of 25% calcium ammonium nitrate, 18% superphosphate and 50% potassium chloride. The major results were as follows: 1. In this dry summer the yields amounted to 0.7 t seed + 5.8 t straw on soil unfertilised for 11 years and 1.8 t seed + 10.5 t straw on plots with maximum NPK supplies. The N uptake amounted to 16-76 kg/ha in the rosette stage, 63-236 kg/ha at flowering and 94-221 kg/ha at harvest in the aboveground yield as a function of NxP nutrition, which was decisive. Of the N adsorbed, 65% was found in the straw and 35% in the seed yield. The roots accumulated only 9-25 kg N by harvest, amounting to some 10% of the aboveground N uptake on average. 2. The specific nutrient requirements of winter rape, i.e, that of 1 t seed + its by-products, were as follows: 130-140 kg N, 110-130 kg K (130-160 kg K2O), 80-100 kg Ca (110-140 kg CaO), 22-28 kg Mg (36-46 kg MgO), 19-24 kg P (44-55 kg P2O5), 9-22 kg Na, 600 g Fe, 300 g Mn, 100 g Zn and 15 g Cu. Due to the low seed yield and the high by-product/seed yield ratio, which was around 6-7 in this experiment, the specific parameters were higher than usual. Specific contents of 100 kg N, 45 kg P2O5, 120 kg K2O, 100 kg CaO and 34 kg MgO can be considered as satisfactory guidelines for the extension service on the basis of previous work under similar conditions. 3. If harvesting is carried out using combines and only the seed yield is removed from the field, I t seed represents a loss of 53 kg N, 16 kg P2O5, 10 kg K2O, 11 kg CaO and 8 kg MgO. The N and P losses should thus be taken into consideration when planning mineral fertilisation, while those of the other nutrients can be ignored. It should be noted that during the period from flowering to harvest the nutrient reserves of the roots rose on average by 20-60% in the case of K, Mg, P, Fe, Mn and Cu, by 100-140% for N, Na and Zn and by 224% for Ca, in contrast to the nutrient losses recorded in the aboveground organs. This concentration of nutrients was especially pronounced on P-deficient soil, being many times greater than that recorded on soil well supplied with P. 4. The mean 4.4 g thousand-seed mass of winter rape was not influenced to any great extent by the treatments. The oil content of the seed declined from 41.8% to 39.67% as the result of N fertilisation. The mean fatty acid composition was in agreement with that characteristic of the variety: 60.1% oleic acid, 24.6% linoleic acid, 9.4% linolenic acid, 4.4% palmitic acid, 1.5% eicosenic acid and 0.8% erucic acid. As the N and P supplies improved there was a significant rise in the linolenic and linoleic acid and a reduction in the oleic acid content of the seed. The oil yield increased from 338 kg/ha in the control to 719 kg/ha in the combined NP fertilisation treatment. Over-fertilisation did not lead to a yield depression. No significant yield surplus was obtained with AL-P2O5 or AL-K2O supplies in excess of 150-200 mg/kg or above an annual N rate of 100 kg/ha, although the oil yield still tended to increase even on over-fertilised soil. 5. The maximum leaching of KCl-soluble NO3-N and SO4-S after 11 years was observed in the 60-200 cm soil layer. Some 30-50% of the fertiliser N and,20-25% of the fertiliser S was demonstrated in soluble form. In the maximum NPK treatments a total of 13.2 t calcium ammonium nitrate, 16.5 t superphosphate and 9.0 t 50% KCl was applied in the course of 11 years, amounting to 38.7 t/ha in all. The total salt content calculated on the basis of electrical resistance was found to be 32 t/ha in the 3 m soil profile, which was of the same order of magnitude as the quantity applied, taking into account the plant uptake (salt balance). The 1-2 m soil layer accumulated the most salt. In this layer the salt content rose to around 0.1%, reflecting the environmental pollution caused by intensive mineral fertilisation.
