Total soluble solids (TSSs) serve as a crucial maturity indicator and quality determinant in apricots, influencing harvest timing and postharvest management decisions. This study develops an advanced framework integrating adaptive boosting (Adaboost) ensemble learning with high-frequency spectral variables selected by uninformative variable elimination (UVE) for the rapid non-destructive detection of fruit quality. Near-infrared (NIR) spectra (1000~2500 nm) were acquired and then preprocessed through robust principal component analysis (ROBPCA) for outlier detection combined with z-score normalization for spectral pretreatment. Subsequent data processes included three steps: (1) 100 continuous runs of UVE identified characteristic wavelengths, which were classified into three levels—high-frequency (≥90 times), medium-frequency (30–90 times), and low-frequency (≤30 times) subsets; (2) the development of the base optimal partial least squares regression (PLSR) models for each wavelength subset; and (3) the execution of adaptive weight optimization through the Adaboost ensemble algorithm. The experimental findings revealed the following: (1) The model established based on high-frequency wavelengths outperformed both full-spectrum model and full-characteristic wavelength model. (2) The optimized UVE-PLS-Adaboost model achieved the peak performance (R = 0.889, RMSEP = 1.267, MAE = 0.994). This research shows that the UVE-Adaboost fusion method enhances model prediction accuracy and generalization ability through multi-dimensional feature optimization and model weight allocation. The proposed framework enables the rapid, non-destructive detection of apricot TSSs and provides a reference for the quality evaluation of other fruits in agricultural applications.

Apricot is a widely cultivated fruit tree of the drupe family, and its sweet/bitter kernel traits are important indicators of the quality and merchantability of apricots. The sweetness/bitterness traits were mainly determined by amygdalin content. However, the lack of high-quality genomes has limited insight into the traits. In this study, a high-quality genome of ‘Xiaobaixing’ was obtained by using single-molecule sequencing and chromosome-conformation capture techniques, with eight chromosomes of 0.21 Gb in length and 52.80% repetitive sequences. A total of 29,157 protein-coding genes were predicted with contigs N50 = 3.56 Mb and scaffold N50 = 26.73 Mb. Construction of phylogenetic trees of 15 species of Rosaceae fruit trees, with ‘Xiaobaixing’ differentiated by 5.3 Ma as the closest relative to ‘Yinxiangbai’. GO functional annotation and KEGG enrichment analysis identified 227 specific gene families to ‘Xiaobaixing’, with 569 expansion-gene families and 1316 contraction-gene families, including the significant expansion of phenylalanine N-monooxygenase and β-glucosidase genes associated with amygdalin synthesis, significant contraction of wild black cherry glucoside β-glucosidase genes, amygdalin β-glucosidase genes, and β-glucosidase genes, and significant enrichment of positively selected genes in the cyanogenic amino acid metabolic pathway. The 88 bHLH genes were identified in the genome of ‘Xiaobaixing’, and ParbHLH66 (rna-Par24659.1) was found to be a key gene for the identification of sweet/bitter kernels of apricots. The amino acid sequence encoded by its gene is highly conserved in the species of Prunus mume, Prunus dulcis, Prunus persica, and Prunus avium and may be participating in the regulation of amygdalin biosynthesis, which provides a theoretical foundation for the molecular identification of sweet/bitter kernels of apricots.

The composition and content of sugar and acid are important indicators of organ development and the quality of horticultural products. B. juncea is an important vegetable in the cruciferous family, with a swollen fleshy stem as the edible organ. Elucidating the characteristics of sugar and acid accumulation during stem expansion of stem mustard and its regulatory mechanism could enhance stem quality. In this study, physiological indicators such as dry matter content, sugar–acid content and related enzyme activities were measured in eight stages of the tumorous stem. The results showed that the sugar and sucrose contents initially increased and then decreased during stem development: Sucrose exhibited a positive correlation with sucrose synthase and sucrose phosphate synthase, while acid content was highly positively correlated with malate dehydrogenase. Further analysis of the dynamic patterns of sugar and acid metabolite contents using metabolomics showed that 1097 metabolites were detected, including 229 organic acids and derivatives, 109 lipids, and other metabolites. Metabolic pathway enrichment analysis showed that metabolites were significantly enriched in organic acids, amino acids, glycolysis/gluconeogenesis, starch, and sucrose metabolism. Analysis of the sugar pathway and the tricarboxylic acid cycle revealed obvious differences in the content and type of metabolites, with most upregulated metabolites in S3 and S4. The expression patterns of enzyme genes associated with the biosynthesis and accumulation of sugar and acid metabolites were found based on differentially expressed genes at different developmental stages, and gene expression levels were verified by qPCR, which showed that the expression patterns of enzyme genes associated with this pathway were highly consistent with the metabolite accumulation. These results suggest that amino acids, sugars, and acids play a critical role in regulating the development of tumorous stems. Results of the present study offer a theoretical basis to improving the quality of mustard by using a metabolomics approach to determine the dynamics of metabolites and related regulatory enzymes during development and the correlation between these levels, which provides novel insights into the potential mechanisms underlying sugar and acid metabolism.

The ‘Diaoganxing’ is the experimental material, with natural grass cover as the control, to compare the effects of 5 different mulching materials. The aim was to identify the most effective mulching type for improving orchard microenvironments and fruit quality. The results demonstrated that waterproof, breathable film and reflective film significantly enhanced orchard microenvironments and fruit quality (p ≤ 0.05). Specifically, the waterproof, breathable film effectively regulated soil temperature and moisture, reducing soil temperature by 4.60% and increasing soil moisture by 17.09% in the 0–60 cm soil layer. Meanwhile, the reflective film optimized light distribution in the mid-lower canopy, increasing light intensity by 161.04–208.71% and reflectance by 2.6–3.3 times. In terms of fruit quality, the reflective film accelerated ripening by 10 d, increased carotenoid content by 15.34%, and achieved a peel color index (CCI) of 6.23. On the other hand, the waterproof breathable film advanced maturation by 7 d and significantly improved vitamin C, soluble sugar, and soluble solids content by 23.26%, 30.77%, and 12.76%, respectively. This study provides a scientific basis for the efficient and high-quality production of apricots in southern Xinjiang through the use of mulching practices.

Apricot (Prunus armeniaca L.) is an important stone fruit in the world that originated in China. The flavor of apricot fruit is mainly determined by its sugar and acid components. Studies have shown that apricot landraces and traditional cultivars can be distinguished by determining the sugar and acid components of the fruit using high-performance liquid chromatography (HPLC) and machine learning approaches. Total soluble sugar, malic acid, and citric acid exhibited no significant differences between the two groups. However, correlation analysis indicated that sugar coordination was enhanced in traditional cultivars compared with landraces, while the organic acid composition was more diverse. Four machine learning models classified them with performance metrics above 0.85. UpSetR analysis identified oxalic acid, tartaric acid, and fructose as key indicators, further improving model performance vs. full features. This study provides a novel strategy based on machine learning to distinguish apricot domestication traits across periods.