Content profiles of mineral elements and bioactive substances and comprehensive evaluation of Wangmo chestnut (Castanea mollissima) kernels

【Objective】This study aimed to clarify the content characteristics of mineral elements and bioactive substances in the kernels of Wangmo chestnuts (Castanea mollissima Blume) and screen Wangmo chestnut resources with high nutritional value for subsequent breeding applications.【Methods】The 20 superior Wangmo chestnut accessions were selected as test materials, and 13 key nutritional indices in their mature fruits were determined. These indices included 9 mineral elements (P, K, Ca, Mg, Fe, Mn, Zn, Cu, and B) as well as 4 bioactive and functional components (total phenolics, vitamin C, tannin, and flavonoids). Correlation analysis was first applied to explore the intrinsic associations between the 13 indices. Then, four complementary evaluation methods—Principal Component Analysis (PCA), weighted summation, entropy evaluation method, and entropy-weighting TOPSIS—were integrated to conduct multi-dimensional analysis and comparative quality assessment of the accessions. Additionally, Kendall’s coefficient of concordance test was employed to quantify the consistency of evaluation results across the three methods, ensuring the reliability of the screening outcomes.【Results】The results indicated that kernel quality varied significantly among the 20 superior accessions, with substantial variations observed in both mineral elements and bioactive components. The coefficient of variation (CV) for mineral elements ranged from 9.32% to 80.19%, while that for bioactive components spanned 4.84% to 52.49%. Specifically, the CVs of Ca, Fe, Mn, Zn, Cu, B, total phenols, vitamin C, and tannin all exceeded 20%, reflecting high variability in these indices across accessions. In contrast, the CVs of P, K, Mg, and total flavonoids were below 20%, indicating relatively stable contents of these components. Regarding mineral elements, WM20 exhibited the highest P content (1 546.315 mg · kg-1 ), which was significantly higher than that of all other accessions (P＜0.05); its Ca content (467.952 mg · kg- 1 ) showed no significant difference from WM28 (459.937 mg · kg-1 ) but was significantly higher than that of the remaining accessions (P＜0.05); WM35 had the highest K content (8031.320 mg · kg- 1 ), which was 44.68% higher than the lowest value (WM2, 5 550.941 mg·kg-1 ), and also maintained high levels of other macro elements (P, Ca, Mg); WM35 also recorded the highest Zn content (27.759 mg · kg- 1 ), whereas WM4 had the lowest (5.927 mg·kg-1 ); WM31 possessed the highest Cu content (8.765 mg·kg-1 ), significantly exceeding that of other accessions (P＜0.05); and WM23 had the highest B content (10.682 mg · kg- 1 ), followed by WM35 (9.371 mg · kg- 1 ). For bioactive components: WM7 had significantly higher contents of total phenols (2.630 mg · g-1 ) and tannin (363.727 nmol· g-1 ) than all other accessions (P＜0.05); in contrast, WM5 had a notably lower total phenol content (0.857 mg · g- 1 ), which was 67.4% lower than that of WM7, and WM31 had a lower tannin content (156.263 nmol· g-1 ), 57.1% lower than WM7. The vitamin C content across accessions ranged from 0.068 to 0.449 mg · g- 1 , with WM28 having the lowest (0.068 mg · g- 1 ), WM40 the highest (0.449 mg · g- 1 ), and WM35 the second highest (0.417 mg · g-1 ). Correlation analysis revealed that vitamin C was highly significantly positively correlated with P and K (P＜0.01) and significantly positively correlated with Mg (P＜0.05); tannin was extremely significantly positively correlated with Fe and total phenols (P＜0.01) but significantly negatively correlated with Cu (P＜0.05); notably, total flavonoids showed no significant correlation with the other 12 nutritional indices. PCA extracted 5 principal components based on eigenvalue＞1, with a cumulative contribution rate of 74.089%, indicating these components explained most of the variability in the original indices. Among them, vitamin C, P, B, Mn, Ca, and Fe were identified as key nutritional indices, as the cumulative variance contribution rate of the first three principal components reached 56.480%. Comprehensive evaluation via PCA highlighted WM35, WM23, WM36, WM20, and WM28 as accessions with superior overall quality. The entropy evaluation method, which quantifies index importance objectively, calculated the weight coefficients of the 13 indices in the order: Cu＞total phenols＞Zn＞B＞vitamin C＞Mg＞Mn＞total flavonoids＞Ca＞K＞Fe＞P＞tannin (tannin was treated as a negative index, while the other 12 were positive indices). This method screened 5 core superior accessions: WM35, WM31, WM3, WM23, and WM40. The entropy- weighted TOPSIS method, which integrates entropy-derived weights to avoid subjective bias, constructed a weighted matrix, calculated the distance to the positive ideal solution (T+ ), distance to the negative ideal solution (T- ), and relative closeness (C); its top 5 superior accessions were consistent with those identified by the entropy evaluation method. Kendall’s coefficient of concordance test showed an extremely strong consistency between the entropy evaluation method and entropy- weighting TOPSIS (concordance coefficient = 0.81, P＜0.05), while PCA exhibited slightly weaker consistency with the two methods (concordance coefficient = 0.57, P＜0.05). For instance, WM6 ranked 8th in PCA but 10th in the entropy evaluation method and 9th in the entropy-weighted TOPSIS method. Notably, the entropy-weighting TOPSIS effectively balanced the importance of different indices and the quality uniformity of individual accessions, demonstrating higher adaptability in comprehensive quality evaluation. The 5 accessions screened by this method all had consistently high contents of mineral elements and bioactive substances.【Conclu-sion】This study clarified the content variation characteristics of mineral elements and bioactive substances in the kernels of superior C. mollissima‘Wangmo’kernels, verified the applicability of the entropy- weighting TOPSIS method for the comprehensive evaluation of superior chestnut kernels, and screened out 5 core superior germplasm resources. These findings provide a solid foundation for the nutritional quality enhancement and efficient utilization of Wangmo chestnut germplasm resources.