Genetic variability and diversity analysis for some agronomic traits of a sweet potato (Ipomoea batatas L.) collection: Insights for breeding superior genotypes
Zakaria Alam,Md Anwar Hossain Khan,Md Iqbal Hossain,Md Rezaul Karim,Hasib-Bin Saif,A A M Mohammad Mustakim,Md Mosharraf Hossain Molla,Md Monirul Islam,Sohela Akhter,Sanjida Akter
DOI: https://doi.org/10.1016/j.heliyon.2024.e38616
IF: 3.776
2024-09-27
Heliyon
Abstract: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.