Genetic diversity in sugarcane (Saccharum spp. L.) from the active germplasm collection of PHILSURIN [Philippine Sugar Research Inst., ] based on coefficient of parentage, agromorphological traits and DNA microsatellite markers

Results showed that a moderate level of genetic diversity was present in the 81 sugarcane cultivars evaluated. Pedigree analysis indicated that a total of 22 ancestral lines formed the foundation of the 73 cultivars studied. Nine major ancestral clones were identified based on frequency of presence and relative genetic contribution. Frequency of presence of the major ancestors was 75.0% while the relative genetic contribution of each line was greater than or equal to 3.0% in the cultivars evaluated. The ancestral lines represent S. officinarum (Benjermazin Hitam, Lahaina, Black Figi and Black Cheribon), S. officinarum X S. sinense (Loethers), S. barberi (Chunnee and Kansar) and S. spontaneum (Java, India). These clones except Kansar are likewise present in 70-79% of introduced cultivars in 71-93% of the VMC cultivars and in all (100%) the SRA cultivars. The breadth of cytoplasmic inheritance was also limited only to three major S. officinarum sources, Benjermazin Hitam, Lahaina and Black Figi. Agromorphological traits showed high Shannon-Weaver diversity indices (0.80) for quantitative traits and moderate (0.5-0.75) to high (0.75) for most of the qualitative traits. Mean Euclidian distance was 52.17 between pairs of cultivars for all possible pair wise combinations and ranged from 11.06 to 131.27. Principal component analysis conducted based on correlation matrix of 16 agromorphological traits resulted to four principal component axes that accounted for 76.22% of total variation. The first principal component that accounted for 29.31% of total variation was mainly attributed to variation in juice quality, yield and stalk diameter traits. Cluster analysis by UPGMA based on Euclidian distances classified the 81 cultivars into one major and four minor clusters. The nine microsatellite markers revealed high gene diversity (PIC) values in the 81 sugarcane varieties. Average gene diversity value was 0.88. SMC 119CG showed the highest gene diversity value that makes it the most discriminating of the nine markers used. Cluster analysis showed that VMC67342, Q50 and VMC74-530 were identical. Another pair, CP43-47 and Q83 were also identical. The nine markers disciminated 94% of the total cultivars surveyed. Cluster analysis of the 81 cultivars resulted to two major and eight minor clusters. Both automated and manual sequencing methods were found efficient in detecting variability in the 48 cultivars studied. SSR bands detected by SSR primers using the manual sequencing method represent about 80% of the total number of alleles resolved in the automated system, with a significant correlation coefficient of r=0.78. Correlation between the genetic distance matrices generated in each method was also significant (r=0.57). Mean PIC values of the four SSR markers were comparable in the automated (0.95) and in the manual (0.91). Correlation values between I-COP and Euclidian distances and between I-COP and SSR-based genetic distance were insignificant. Euclidian distance and SSR-based genetic distance were significantly correlated but the correlation coefficient was low (r=0.12). The relationship between Eiclidian distance and SSR-based genetic distance was not linear and appeared to be "triangular". Recommendations were presented.