基于多元统计分析的新疆西梅品质多维度综合评价

丁真真; 王学庆; 苏婷婷; 王莉; 王玉涛

doi:10.13386/j.issn1002-0306.2026020003

基于多元统计分析的新疆西梅品质多维度综合评价

Multidimensional Comprehensive Evaluation of Xinjiang Prune Quality Based on Multivariate Statistical Analysis

摘要

摘要: 为明确南疆主栽西梅品种的品质特征与加工适宜性差异，本研究系统评价了新疆南疆四大主栽品种（‘法兰西’、‘红喜梅’、‘小西梅’和‘蓝西梅’）7个产地西梅样品的加工适宜性。通过测定外观、质构、理化、活性成分及挥发性风味物质，构建了多维品质指标体系，并采用聚类分析与正交偏最小二乘判别分析（orthogonal partial least squares discriminant analysis，OPLS-DA）对样品进行分类。结果表明，不同西梅品种在花色苷、多酚及可溶性固形物等关键品质指标上存在显著差异。综合品质分析显示，各样品呈现差异化的品质特征：‘蓝西梅’（喀）以高硬度、高出汁率为突出特性；‘蓝西梅’（泽）则表现为高酸与高花色苷含量；‘法兰西’（伽）的可溶性固形物、多酚含量及抗氧化活性显著优于其他样品；‘小西梅’（伽）与‘法兰西’（阿）则以高糖酸比和丰富的特征香气成分为主要优势。所构建的OPLS-DA模型判别性能优异（R²X=0.998，Q²=0.994），并筛选出果型指数、可溶性固形物（total soluble solids，TSS）、多酚、2-己烯醛、DPPH·清除率、黄酮、波数3300 cm⁻¹（O-H伸缩振动）处的红外吸收峰、ABTS⁺·清除率、月桂醛、·OH清除率等14个变量重要性投影（variable importance in projection，VIP）值大于1.0的关键品质标志物。本研究可为西梅专用品种选育、智能化分级及定向加工提供理论依据与技术支撑。

Abstract: To clarify the quality characteristics and processing suitability differences of major cultivated plum varieties in Southern Xinjiang, the processing suitability of Prunus domestica samples collected from seven production regions across four major cultivated varieties (‘Francistown’, ‘Hong Ximei’, ‘Xiao Ximei’, and ‘Blue Plum’) in Southern Xinjiang, was systematically evaluated in this study. A multidimensional quality evaluation system was established by measuring appearance, texture, physicochemical properties, bioactive components, and volatile flavor compounds. Cluster analysis and orthogonal partial least squares discriminant analysis (OPLS-DA) were employed to classify the samples. The results showed significant differences in key quality indicators such as anthocyanins, total polyphenols, and total soluble solids (TSS) among different cultivars. Comprehensive quality analysis revealed distinct quality characteristics among the samples: ‘Blue Plum’ from Kashi exhibited notably high firmness and juice yield; ‘Blue Plum’ from Zepu was characterized by high acidity and anthocyanin content; ‘Francistown’ from Jiashi displayed significantly higher sugar content, polyphenol content, and antioxidant activity compared with the other samples, while ‘Francistown’ from Jiashi and ‘Francistown’ from Aksu were distinguished by their high sugar-acid ratio and abundant characteristic volatile aroma compounds. The constructed OPLS-DA model exhibited excellent discriminative performance, with an R²X of 0.998 and a Q² of 0.994. A total of 14 key quality markers with variable importance in projection (VIP) values greater than 1.0 were identified, including fruit shape index, TSS, polyphenols, 2-hexenal, DPPH· radical scavenging activity, flavonoids, infrared absorption peak at the wavenumber of 3300 cm⁻¹ (O-H stretching vibration), ABTS· radical scavenging activity, lauric aldehyde, and hydroxyl radical scavenging activity. This study provides a theoretical basis and technical support for selective breeding, intelligent grading, and targeted processing strategies in the plum industry.