Alam Zakaria, Khan Md Anwar Hossain, Hossain Md Iqbal, Karim Md Rezaul, Saif Hasib-Bin, Mustakim A A M Mohammad, Molla Md Mosharraf Hossain, Islam Md Monirul, Akhter Sohela, Akter Sanjida
Tuber Crops Research Centre, Bangladesh Agricultural Research Institute (BARI), Gazipur, 1701, Bangladesh.
Horticulture Research Centre, Bangladesh Agricultural Research Institute (BARI), Gazipur, 1701, Bangladesh.
Heliyon. 2024 Sep 27;10(19):e38616. doi: 10.1016/j.heliyon.2024.e38616. eCollection 2024 Oct 15.
A study was conducted during the 2017 cropping season to assess the genetic variability of selected agronomic traits in sweet potato. A total of 355 sweet potato genotypes (351 from different countries of the world and four check varieties) were evaluated at Tuber Crops Research Sub Centre, Bangladesh Agricultural Research Institution, Bogura, Bangladesh. The experiment was conducted using an augmented randomized complete block design. Findings revealed positive correlations among five agronomic traits, namely the weight of marketable storage root per plant (MRW), weight of storage root per plant (RW), fresh weight of foliage per plant (FW), number of marketable storage root per plant (MRN) and number of storage root per plant (RN) suggesting their suitability for phenotypic selection. There was a positive skewness in most studied traits (FW, MRW, MRW and RW), indicating the presence of a few genotypes with exceptionally high values for those traits. By employing principal component analysis, the most influential traits in sweet potato genotypes were identified for first two components (Dim) showing significant contributions to Dim-1 and Dim-2 collectively accounting for 84.7 % of the total variance. Through cluster analysis, two main clusters (Cluster I and Cluster II) were identified among the studied genotypes, each characterized by distinct trait patterns. Cluster I demonstrated superior performance in all traits, while Cluster II exhibited lower values for all traits. Further sub-clustering within Cluster II (sub-clusters IIA and IIB) revealed additional variations of the collection. The Euclidean distance for inter-cluster (ranging from 7.94 to 11.39) and intra-cluster (ranging from 3.17 to 10.38) were determined utilizing the complete distance method. In restricted maximum likelihood (REML) model, all the traits were significantly influenced by genetic effects, with broad-sense heritability ranging from 32 % to 70 %. The REML displayed high selection accuracy, with values exceeding 91 % for studied traits. The Multi-Trait Genotype-Ideotype Distance Index (MGIDI) identified 18 genotypes, with Moz1.15 emerging as the best-performing genotype under a 5 % selection pressure. Among the selected genotypes, trait uniqueness was highest for FW (54 %), followed by RN (50 %), MRN (18 %), MRW (14 %), and RW (5 %). The selection gain for these top-ranked genotypes varied between 23.2 % and 69.2 %. Findings established the existence of useful genetic variability for selected agronomic traits of sweet potato that could be exploited for breeding, genetics study and conservation.
2017年种植季开展了一项研究，以评估甘薯某些农艺性状的遗传变异性。在孟加拉国博古拉的孟加拉农业研究机构块根作物研究分中心，对总共355个甘薯基因型（351个来自世界不同国家，4个为对照品种）进行了评估。试验采用增广随机完全区组设计。研究结果显示，五个农艺性状之间存在正相关，即单株可售贮藏根重量（MRW）、单株贮藏根重量（RW）、单株叶鲜重（FW）、单株可售贮藏根数量（MRN）和单株贮藏根数量（RN），这表明它们适合进行表型选择。大多数研究性状（FW、MRW、MRW和RW）呈正偏态，表明存在少数这些性状值极高的基因型。通过主成分分析，确定了甘薯基因型中对前两个成分（Dim）影响最大的性状，这两个成分对Dim - 1和Dim - 2的贡献率合计占总方差的84.7%。通过聚类分析，在所研究的基因型中识别出两个主要聚类（聚类I和聚类II），每个聚类具有不同的性状模式。聚类I在所有性状上表现优异，而聚类II在所有性状上的值较低。聚类II内的进一步亚聚类（亚聚类IIA和IIB）揭示了该群体的更多变异。利用完全距离法确定了聚类间（范围为7.94至11.39）和聚类内（范围为3.17至10.38）的欧氏距离。在限制最大似然（REML）模型中，所有性状均受到遗传效应的显著影响，广义遗传力范围为32%至70%。REML显示出较高的选择准确性，所研究性状的值超过91%。多性状基因型 - 理想型距离指数（MGIDI）识别出18个基因型，在5%的选择压力下，Moz1.15成为表现最佳的基因型。在所选择的基因型中，FW的性状独特性最高（54%），其次是RN（50%）、MRN（18%）、MRW（14%）和RW（5%）。这些排名靠前的基因型的选择增益在23.2%至69.2%之间。研究结果证实了甘薯某些农艺性状存在有用的遗传变异性，可用于育种、遗传学研究和保护。
