A lysimeter experiment was carried out over a period of three years on a Ra
mann's sandy, loamy forest soil (CaCO3 2.1 %, sand 1.3 %, poorly supplied w
ith phosphorus and potassium). The lysimeters had a soil volume of m(3) and
a growing surface of 4 m(2). Basic fertilisation with lao kg/ha P2O5 and 1
20 kg/ha K2O was applied each year. N was applied at rates of 0, 40, 80 and
120 kg/ha in 1986 and 0, 100, 150 and 200 kg/ha in 1987 and 1988 in the fo
rm of NH4NO3. Soya was sown without inoculation at a plant density of 500 0
00 plants/ha in a total of 4 treatments x 3 replications = 12 lysimeters. T
he mass of the individual plant parts, the numbers of flowers, pods and see
ds per plant, the thousand seed mass and the leaf area per plant were deter
mined. The main results are summarised below.
1. As the N supplies improved the leaf area increased 2.2 times by the end
of flowering in 1986, reaching a LAI value of 9 m(2)/plant. At the same tim
e the seed yield and the total biomass production were also almost doubled
compared to the control.
2. At satisfactory N supply levels the number of improductive flowers dropp
ed to almost half, averaged over the years. Both deficient and excessive N
supplies increased the number of improductive flowers, thus reducing the nu
mber of pods per plant, since the total number of flowers was only slightly
modified as the result of N fertilisation. Due to the balancing out of yie
ld components, a larger number of pods per plant often leads to a lower num
ber of seeds per pod. The yield surplus compared to the control was associa
ted with the cumulative effect of a larger number of pods and seeds per pla
nt, and higher thousand seed mass.
3. The ratio of the seed yield to the total biomass production (harvest ind
ex) averaged 20-30 %, while the ratio of N incorporated in the seeds (nitro
gen harvest index) fluctuated between 50 and 70 % over the years. Although
the seeds have an extremely high N content, the residual by-products are al
so able to improve the N balance of the soil, thus reducing the need for N
fertilisers.
4. A maximum seed yield of 5.4 t/ha was achieved in 1987 in the 150 kg/ha N
treatment, with the following yield components: number of pods per plant 2
5, average seed number per pod 1.7, number of seeds per plans 44, thousand
seed mass 241 g. The outstanding seed yield could be attributed to a greate
r number of pods per plant and to the extremely high thousand seed mass. Th
ese data confirm the yield potential of soya under Hungarian conditions.