Principal component analysis for assessment of variability in phenological and morphological traits in French bean (Phaseolus vulgaris L)

Abstract

Forty four french bean (Phaseolus vulgaris L.) genotypes were evaluated for phenological, morphological, pod and seed traits in an augmented block design and variability was assessed through principal component analysis. PCA concentrated variability in first six principal components. The total variance explained with the first six PC’s was 80.366 %.  Latent roots (Eigen values) for significant PCs ranged from 3.390 (PC1) to 1.056 (PC5). The first two PC’s that were used for constructing biplot graphs explained 41.259 %. The first PC contributed 22.601% of total variation mainly contributed by SYPP, NPPP, followed by PYPP, PGR and SPP.  The second component explained 18.658 % of variation contributed largely by DF, followed by DPF, DM, PGR  and PL, while as the third, fourth, fifth and sixth component explained 13.615, 10.390, 8.064 and 7.038 % of variation respectively. In the present study, seven variables including SYPP, PYPP, PGR, NPPP, DF, DPF, and DM contributed above the expected average to the variability in PC1 and PC2. Based on the factor loading graph, seed yield per plant is strongly correlated with number of pods and seeds per pod. Similarly pod yield per plant is strongly correlated with pod growth rate, pod length and number of pods per plant.  Seed yield per plant is negatively correlated with seed length and plant height while as no correlation with days to maturity can be visualized. The genotypic profiles in biplot revealed that genotypes WB-129, WB-371, WB-1187, WB-642 and WB-1518 have  high pod yield per plant  as all of them have high pod growth rate, high  pod number as well as higher pod length.