Analysis of Plant Ideotype and Yield in Hybrid Maize under Varied Population Densities

Optimizing maize yield within limited productive land is a critical challenge due to the rapid conversion and reduction of land suitable for maize cultivation. The objective of this study was to evaluate the leaf orientation ideotype of maize genotypes across different population densities and to assess its implications for agronomic traits such as light interception, canopy area, and grain yield. In addition, machine learning-based modeling has also been applied to predict the yield under various population settings. The study was conducted at the Bajeng Experimental Station in Indonesia, using a split-plot design. The main plots encompassed 11 genotypes along with two commercial test varieties characterized by an upright leaf type, while the subplots examined three population levels i.e. normal density (71,428 plants ha-1 ), PHGLXP (81,632 SODQWV KDၱï) DQG KLJK GHQVLW\ (95,238 SODQWV KDၱï). TKH results revealed an 11% increase in grain yield under medium density compared to standard and high-density populations. Hybrid genotypes affect yield variability, emphasizing the need for hybrids that areadapted to dense populations. Canopy structure, particularly leaf angle and curvature, influences light interception and photosynthesis, thereby enhancing yields in dense populations. Hybrids with erect leaves, smaller angles, reduced area, and larger stem diameter were best in dense settings, as demonstrated by H06 and H13, promising increased yield under dense planting systems. Machine learning assessment indicated that random forest and SVM outperform multiple linear regression in maize yield under varying population architectures, achieving R2 values of 0.69 and 0.62, respectively. Keywords: ideotype; maize hybrid; population density; erect leaves; machine learning