Selection for high-yielding sugarcane ideotypes and multi-trait selection via MGIDI across different harvesting periods improves sugar quality and yield

Evaluating sugar-related traits at different harvesting times is crucial for determining the best times to harvest in order to maximize yield and minimize losses from over-ripening. This study investigated the genetic and agronomic potential of 30 sugarcane (Saccharum officinarum L.) genotypes across multiple growth stages (395–480 days after sowing, DAS) to optimize sugar yield (SY) and processing quality through advanced selection methodologies. By utilizing the Multi-Trait Genotype-Ideotype Distance Index (MGIDI) and Pesek–Baker Index (PBI), elite genotypes such as G09 and G07 were identified, showing superior sugar content (brix = 16.6–17.8%; commercial cane sugar, CCS = 9.13–10%) and stable SY (12.68–13.41 t ha⁻¹). Factor analysis revealed three latent drivers of productivity, explaining 88% of total variability. High heritability (> 80% for CCS and total recoverable sugar, TRS) and high genetic advance (> 15%) were recorded for sugar-related traits, particularly at the mid-late stages of growth (454–480 DAS). Regression analysis identified stem yield, polarization, and stalk diameter as primary contributors to SY, explaining 79% of the variance. Temporal trends confirmed progressive sucrose accumulation peaking at 480 DAS, with high genotypic coefficient of variation (GCV) in RS and CCS observed in the early stage of growth. The integration of MGIDI and PBI prioritized purity as an industrial-critical trait (PBI = 2.75). While G23 and G11 performed poorly, the genotypes G03 and G05 had high SY (> 13 t ha⁻¹). In conclusion, this study emphasizes the importance of using a trait-weighted- based selection index, temporal trait evaluation and multi-index selection in breeding programs of sugarcane to develop cultivars with synchronized high sugar content and yield stability.