不同锥栗农家种种仁中9种矿质元素含量的因子分析与聚类分析

Factor Analysis and Cluster Analysis of Contents of 9 Mineral Elements in Seed Kernels of Castanea henryi from Different Varieties

为探明锥栗种仁中矿质元素含量特征规律,以30种主要锥栗农家种种仁为试材,采用自动间断分析仪与原子吸收光谱法测定N、P、K、Ca、Mg、Fe、Mn、Zn和Cu 9种矿质元素指标,运用相关性分析、因子分析和聚类分析对其矿质元素含量进行分析。结果表明,Kolmogorov-Smirnov检验表明9种矿质元素含量数据服从正态分布,9种矿质元素平均含量顺序为N(6 873.74 mg/kg)>K(4 402.32 mg/kg)>P(1 619.16 mg/kg)>Ca(471.18 mg/kg)>Mg(394.59 mg/kg)>Mn(115.80 mg/kg)>Fe(16.22 mg/kg)>Zn(8.68 mg/kg)>Cu(7.59 mg/kg),大量元素中N、K、P与Ca、Mg存在显著差异(P<0.05),微量元素中Mn与Fe、Zn、Cu差异显著(P<0.05),变异系数范围为8.29%～54.43%;相关性分析表明各元素之间存在着复杂的关联性;因子分析结果表明,N、P、Mn、K、Cu和Fe是锥栗的特征元素,提取的6个公因子累计方差贡献率为90.572%,第1公因子方差贡献率为22.400%,主要综合了N、P和Mn 3种元素的信息,第2公因子方差贡献率为15.572%,主要综合了K和Cu 2种元素的信息,第3公因子方差贡献率为14.701%,与Fe有关,第4公因子方差贡献率为14.614%,与Mg有关,第5公因子方差贡献率为11.936%,与Ca有关,第6公因子方差贡献率为11.349%,与Zn有关,综合得分排名前5名依次为蔓榛、长芒仔、中尖嘴、材榛和小尖嘴;从元素含量角度进行聚类分析,30种锥栗农家种可分为6类。本研究结果可为进一步开展锥栗营养功能评价、锥栗育种亲本选择和锥栗食品开发等提供基本参考数据。

In order to elucidate the characteristics of mineral elements in the seed kernels of Chinese chinquapin (Castanea henryi), the contents of nine mineral elements including N, P, K, Ca, Mg, Fe, Mn, Zn and Cu in 30 major varieties were determined using a discontinuous autoanalyzer and atom absorption spectrometry and they were analyzed by correlation analysis, factor analysis and cluster analysis. The results were showed that the contents of all 9 mineral elements obeyed normal distribution by Kolmogorov-Smirnov test. The average contents followed the descending order of N (6 873.74 mg/kg) > K (4 402.32 mg/kg) > P (1 619.16 mg/kg) > Ca (471.18 mg/kg)> Mg (394.59 mg/kg) > Mn (115.80 mg/kg) > Fe (16.22 mg/kg) > Zn (8.68 mg/kg) > Cu (7.59 mg/kg). As far as the macroelements were concerned, N, K and P were significantly different from Ca and Mg (P < 0.05); as for the microelements, Mn was significantly different from Fe, Zn and Cu (P < 0.05). The coefficients of variation varied from 8.29% (Zn) to 54.43% (Fe). Correlation analysis showed that there was a complex correlation among these elements. The results of factor analysis showed that N, P, Mn, K, Cu and Fe were the characteristic elements of Chinese chinquapin, and that the cumulative variance contribution rate of six common factors extracted was 90.572%. The cumulative variance contribution rate of the first common factor (F1) was 22.400%, representing N, P and Mn. The cumulative variance contribution rate of the second common factor (F2) was 15.572%, representing K and Cu. The cumulative variance contribution rate of the third common factor (F3) was 14.701%, representing Fe. The cumulative variance contribution rate of the fourth common factor (F4) was 14.614%, representing Mg. The cumulative variance contribution rate of the fifth common factor (F5) was 11.936%, representing Ca. The cumulative variance contribution rate of the sixth common factor (F6) was 11.349%, representing Zn. The top five varieties with the highest synthetic scores were ‘Manzhen’, ‘Changmangzi’, ‘Zhongjianzui’, ‘Caizhen’, ‘Xiaojianzui’. The 30 Chinese chinquapin varieties were classified into 6 groups by cluster analysis. The results of this study can provide basic data for evaluating nutritional properties of Chinese chinquapin, breeding new varieties and developing Chinese chinquapin-based foods.