Field Crop 2025, Vol.8, No.3, 102-112 http://cropscipublisher.com/index.php/fc 108 On the other hand, high-throughput phenotyping and artificial intelligence can also be used to achieve intelligent selection. For example, drone hyperspectral imaging is used to monitor the growth, physiology, and disease status of field sugarcane, and these data are combined with flavor quality determination. Machine learning models are used to predict the comprehensive breeding value of each strain to assist in screening (Mahadevaiah et al., 2021). At present, studies have used machine learning to explore sugarcane rust resistance genetic regions and predict the impact of sugar reduction policies on sugarcane planting (Aono et al., 2020; Thow et al., 2021). Similar methods can also be used for the trade-off optimization of multiple traits in fresh sugarcane. In addition, decision support systems (DSS) can also be applied. A database of multi-trait data from breeding trials for many years is established, and decision support software is developed. After entering the target weight, candidate strains that meet the requirements can be automatically recommended. 5.3 Phased multi-objective aggregation strategy Multi-objective breeding does not require that all traits be improved in the same generation. In actual breeding, a phased aggregation strategy is often used. The first stage focuses on basic agronomic traits, such as yield and disease resistance screening, and eliminates obviously poor materials; the second stage introduces and selects quality traits while retaining high-quality agronomic traits. Specifically, yield and stress resistance can be mainly examined in large groups of early generations, such as inoculating and identifying smut resistance at the seedling stage, investigating rust resistance in the field, and measuring sugar content at the same time, leaving high-yield, high-sugar and disease-resistant materials. Then, the flavor quality of the selected materials is measured and evaluated in the mid-generation, including sensory tasting and physical and chemical analysis (de Queiroz Bomdespacho et al., 2021). For those with poor flavor, they can be discarded even if the yield is high, so as to ensure that the remaining lines have a certain level in both yield and flavor. Finally, in the regional trial stage, adaptability and commerciality are comprehensively examined, and the final selection is made based on the aforementioned comprehensive index. 5.4 Molecular design and gene editing-assisted breeding With the deepening of understanding of sugarcane functional genome, molecular design breeding can be carried out, that is, genotype design and modification based on the genetic mechanism of target traits (Govindakurup and Mohanraj, 2024). For multi-target traits, the ideal genotype may involve the optimal combination of multiple genes. With the help of genotype data and system biology models, the effects of different allele combinations on phenotypes can be simulated to guide hybridization. For example, genome prediction can be used to select complementary parent hybrids to make the offspring have both high sugar and high resistance allele combinations (Mahadevaiah et al., 2021). Gene editing provides a means to directly modify key genes. For example, for acid metabolism genes that affect flavor and susceptibility genes that affect disease resistance, CRISPR/Cas9 can be used to knock out or edit them, thereby improving sweetness or resistance without introducing exogenous fragments. Recent studies have edited the sugarcane CHI gene and found that it can change the plant's immune response and improve disease resistance (Chen et al., 2025). 6 Case Analysis of Typical Excellent Fresh Sugarcane Varieties 6.1 Case 1: comprehensive trait balanced variety "Guitang 42" Guitang 42 is a sugarcane variety bred by the Sugarcane Research Institute of Guangxi Academy of Agricultural Sciences. As a star representative of Guangxi's "Gui" series of varieties, it passed the Guangxi variety approval around 2014. The reason why this variety is typical is that it achieves a balance between high yield, high sugar, disease resistance and wide adaptability, and is known as an excellent variety with "multi-resistance and high sugar". GT42 has tall plants, strong stems, strong tillering ability and perennial rooting, and has early maturity characteristics. Generally, it can reach sugar production maturity in November. In terms of disease resistance, GT42 shows strong resistance to major sugarcane diseases. For example, it is reported that GT42 carries the rust resistance Bru1 gene, so it is immune to brown rust; at the same time, it also has a medium resistance level to black smut and is resistant to sugarcane mosaic (Li et al., 2023). In addition, GT42 has an upright plant shape, strong stems, and outstanding lodging resistance (Figure 3). Li et al. (2023) pointed out that GT42 was obtained through conventional hybrid breeding, and has strong comprehensive excellent traits such as strong lodging resistance and early maturity, and has become one of the main varieties in Guangxi.
RkJQdWJsaXNoZXIy MjQ4ODYzNA==