Genotype x Nitrogen x Environment Interaction and Grain Yield Determinants in Inbred Lines and Derived Hybrids.

The development of maize (Zea mays L.) genotypes with good performance in low nitrogen (N) environments pursues reducing corn raising costs and allowing the practice of sustainable agriculture in areas affected by N deficiency. The performance of single cross hybrids and their corresponding inbred parents, selected on the basis of differences in features regarding N use in previous studies, were analyzed across a group of four climatic and soil contrasting environments (E). Six inbred lines and all possible two-way hybrids were used (complete diallel experiment). The experimental layout was a factorial design of N availability as the main plot (N0: no fertilizer, N1: 200 kg N ha^-1) and Genotype (G) in the subplots, with three replicates. Treatments affected (P<0.001) grain yield (GY) and its components, and a strong (P<0.001) E x G x N interaction was determined for all studied variables. For the dataset, the variation in GY was mostly explained (88%) by kernel number (KN) per m² (range between 827 and 4735 for hybrids and 256 to 3000 for inbred lines). Principal component analysis also rendered high associations among GY and its components. The first principal component explained up to 82% of the total variance and allowed a separation of treatments based on the impact on KN, and consequently on GY. The second principal component accounted for variations in KW. At each environment, the N1 treatment had the highest KN and GY values. Nevertheless, differences among N treatments were greater in irrigated environments than in the rainfed ones. Considering the uppermost 20% (6 highest yielding hybrids) and the lowermost 20% (6 poorest yielding hybrids) at each environment, three groups of hybrids could be distinguished for all eight E × N conditions. One inbred line stands out as not having any derived hybrid ranked in the poorest grain yielding group.