Evaluation of the effect of plant density on maize production by means of risk analysis

Within the framework of a 4-year polyfactorial experiment set up on chernoz em soil with lime deposits at the experimental station of the Department of Crop Production and agronomy of Debrecen University of Agricultural Scienc es, studies were made to determine the most favourable plant density (N1=60 , N2=70, N3=80, N4=90 thousand plants/ha) in irrigated and non-irrigated tr eatments using the fertiliser level (120 kg N + 90 kg P2O5 + 106 kg K2O/ha) and soil cultivation method (autumn ploughing) found in earlier experiment s to give the best results (Drimba, 1977). The aim was to choose a plant de nsity which was favourable as regards both the size of yield and the extent of risk. The criteria applied were stochastic dominance (SD) and the expected value variance (E-V). The first of these was used to rank the plant density treat ments by determining the plant density which ensured the greatest yield at each projected yield level. The second was used to plot an efficient E-V cu rve where the plant density treatment ratios associated with each point cou ld be employed to determine the smallest risk (V) for each yield (E) value. In the case of non-irrigated treatments the SD analyses indicated that as t he projected yield level rose the desired yield could be achieved with the greatest probability at plant density N1 over the lowest 1/3 of the curve. at N2 over the middle 1/3 and at N3 in the uppermost interval. The second order SD criteria suggested that N1 and N2 dominated the other t wo plant densities, but no difference could be demonstrated between N1 and N2 even using third order SD. If the yield risk is considered, NI or N2 sho uld be applied to achieve the greatest yields (in the ratios which can be r ead from the table). In the case of irrigated treatments the SD analyses showed that the likelih ood of achieving a yield equal to or exceeding the desired value was best e nsured by N2 in the lower 2/3 of the yield range and by N3 above this level . According to the second order SD criteria, the top rankings were achieved by N2 and N3, in this order. From the point of view of yield risk N2 could definitely be recommended for the achievement of high yields, It can be concluded from the above results and from those of previous exper iments that to achieve the desired high maize yields at minimum risk a fert iliser level of 120 kg N + 90 kg P2O5 + 106 kg K2O/ha with autumn ploughing and a density of 70 thousand plants/ha can be recommended in both non-irri gated and irrigated treatments under the conditions obtaining in the presen t experiments, Precise information can be found in these studies for concre te cases, e.g. the reliable achievement of a given yield level in irrigated treatments.