Table_1_Phenotypic evolution in durum wheat (Triticum durum Desf.) based on SNPs, morphological traits, UPOV descriptors and kernel-related traits.xlsx

Durum wheat is a worldwide staple crop cultivated mainly in the Mediterranean basin. Progress in durum wheat breeding requires the exploitation of genetic variation among the gene pool enclosed in landraces, old cultivars and modern cultivars. The aim of this study was to provide a more comprehensive view of the genetic architecture evolution among 123 durum wheat accessions (41 landraces, 41 old cultivars and 41 modern cultivars), grown in replicated randomized complete block in two areas, Metaponto (Basilicata) and Foggia (Apulia), using the Illumina iSelect 15K wheat SNP array and 33 plant and kernel traits including the International Union for the Protection of new Varieties of Plants (UPOV) descriptors. Through DAPC and Bayesian population structure five groups were identified according to type of material data and reflecting the genetic basis and breeding strategies involved in their development. Phenotypic and genotypic coefficient of variation were low for kernel width (6.43%) and for grain protein content (1.03%). Highly significant differences between environments, genotypes and GEI (Genotype x Environment Interaction) were detected by mixed ANOVAs for agro-morphological-quality traits. Number of kernels per spike (h2 = 0.02) and grain protein content (h2 = 0.03) were not a heritability character and highly influenced by the environment. Nested ANOVAs revealed highly significant differences between DAPC clusters within environments for all traits except kernel roundness. Ten UPOV traits showed significant diversity for their frequencies in the two environments. By PCAmix multivariate analysis, plant height, heading time, spike length, weight of kernels per spike, thousand kernel weight, and the seed related traits had heavy weight on the differentiation of the groups, while UPOV traits discriminated moderately or to a little extent. The data collected in this study provide useful resources to facilitate management and use of wheat genetic diversity that has been lost due to selection in the last decades.

硬粒小麦（Durum wheat）是全球主粮作物，主要种植于地中海盆地。硬粒小麦育种工作的推进，有赖于开发地方品种、古老品种与现代品种所构成的基因库中的遗传变异。本研究旨在全面解析123份硬粒小麦种质（其中地方品种、古老品种与现代品种各41份）的遗传结构演化历程：试验采用重复完全随机区组设计，在巴西利卡塔大区的梅庞托（Metaponto）与普利亚大区的福贾（Foggia）两个试点开展，检测时使用Illumina iSelect 15K小麦单核苷酸多态性（Single Nucleotide Polymorphism，SNP）阵列，并测定了33项植株与籽粒性状，其中包含植物新品种保护国际联盟（International Union for the Protection of New Varieties of Plants，UPOV）描述符。通过主成分判别分析（Discriminant Analysis of Principal Components，DAPC）与贝叶斯群体结构分析，研究人员根据材料类型共鉴定出5个类群，该分类结果反映了各类群的遗传基础与育种选育策略。籽粒宽度（6.43%）与籽粒蛋白质含量（1.03%）的表型变异系数与遗传变异系数均较低。针对农业形态与品质性状的混合方差分析（Analysis of Variance，ANOVA）结果显示，试点环境、基因型及其基因型×环境互作（Genotype × Environment Interaction，GEI）均存在极显著差异。每穗粒数（狭义遗传力h²=0.02）与籽粒蛋白质含量（h²=0.03）遗传力极低，极易受环境因素影响。嵌套方差分析结果显示，除籽粒圆度外，各试点内不同DAPC类群间的其余所有性状均存在极显著差异。10项UPOV性状在两个试点中的性状频率分布存在显著多样性差异。通过PCAmix混合主成分分析，株高、抽穗期、穗长、每穗粒重、千粒重以及籽粒相关性状对类群分化的贡献权重极高，而UPOV性状的区分能力则中等偏弱。本研究采集的数据可为科研人员管理与利用近数十年间因人工选育而丢失的小麦遗传多样性提供宝贵的资源支持。