Phenotypic Diversity Analysis, Comprehensive Evaluation, and GWAS of Protein and Sucrose Content in Guizhou Local Peanut Germplasm Based on Three Consecutive Years of Phenotypic Data
Qinglin Rao, Liangqiang Cheng, Min Jiang, Jianwei Lv, Tinghui Hu, Jinhua Wang, Jianqin Zhang, Jun WangBackground/Objectives: Peanut landraces (Arachis hypogaea L.) from Guizhou, a karst region of southwestern China, represent an underexploited breeding resource whose phenotypic architecture and marker associations remain insufficiently characterized. Methods: We evaluated 300 local accessions for 30 morphological, agronomic, and quality traits using three-year mean values and analyzed the data using phenotypic diversity statistics, Pearson correlation, stepwise regression, principal component analysis (PCA), hierarchical clustering, and genome-wide association study (GWAS). Results: The collection exhibited substantial phenotypic variation, with Shannon-Wiener indices ranging from 0.54 for leaf color to 2.09 for the equivalent yield per hectare (fresh weight). The Shannon-Wiener index for seedling uniformity was 0.69 in both Liupanshui and Tongren, indicating comparable phenotypic diversity for this trait rather than a direct measure of seedling uniformity. Guiyang accessions had the highest Shannon-Wiener index for sucrose content (1.99). Among 465 trait pairs, 113 showed significant positive correlations and 76 showed significant negative correlations. Stepwise regression retained 100-kernel weight, leaf color, leaf shape, 100-pod weight, main stem height, length of the first branches, and pod number per plant as the main statistical predictors of quality traits. PCA retained eight principal components with eigenvalues above one, which together explained 75.77% of the total variance. A total of 137 accessions had composite scores ≥ 0, and the top ten, including G23212 and G23172, were selected as candidate elite materials. Cluster analysis divided the accessions into four groups and identified 23 representative types. At a suggestive threshold of −log10(P) > 3, the GWAS detected 33 protein- and 27 sucrose-associated SNPs; these loci should be regarded as candidate associations pending validation. Conclusions: Collectively, these results indicate that Guizhou peanut landraces harbor substantial, geographically differentiated phenotypic variation and provide candidate materials and loci for further germplasm evaluation and marker validation.