Multidimensional scaling for divergence analysis in pepper

Abstract

Pepper plants of the genus Capsicum present wide genetic variability for phenotypical characters. The joint analysis of quantitative and qualitative characters is a more complete indicator to identify the existing variability in populations of plants. This work aimed to evaluate the genetic divergence within F3 populations of Capsicum annuum L. through the multidimensional scaling based on the joint analysis of quantitative and qualitative data. The work was developed in at the Center of Agrarian Sciences of the Federal University of Paraíba, state of Paraíba, Brazil. Eight populations of an F3 generation were employed, performing the characterization of 36 qualitative and quantitative characteristics. The experimental design was in randomized blocks, with eight treatments and 50 replications. The non-metric multidimensional scaling (nMDS) was performed based on Gower’s dissimilarity matrixes. Dispersion diagrams were built to present the nMDS solutions in the bidimensional space. The nMDS mapping was efficient on estimating divergence within populations with Kruskal’s Stress below 20%. The internal uniformity presented by populations P-4, P-5, P-6, P-7, and P-8 is an indicator that the genotypes of these populations are suitable to be used in future genetic breeding programs of ornamental peppers. Conversely, populations P-1, P-2, and P-3 presented the greatest dispersion, characterizing phenotypical variability. The multidimensional scaling was efficient to estimate the genetic divergence among and within F3 populations of ornamental pepper