Caracterización fenotípica y genética de clones promisorios de cacao (Theobroma cacao L.) con análisis multivariante procrustes.

Abstract

Similarities were determined between 26 Criollo-type cacao clones and 4 referential control clones from other genetic groups, with respect to the simultaneous analysis of morphological and molecular characterization variables. Data were collected for 38 morphological variables measured on pods, seeds, leaves and flowers, and for 20 microsatellite molecular markers. From each morphological matrix, a matrix of population distances between clones was obtained, and a similarity matrix was obtained from the molecular matrix. From each distance/similarity matrix the principal coordinates were obtained, each matrix of principal coordinates being a configuration. Based on these, with Generalized Procrustes Analysis, a technique that allows reach consensus on these data, a consensus was obtained between morphological configurations, whose result was compared with another methodology applied on the same data, and finally a consensus between the molecular configuration and the consensus of the morphological configurations (Software: Excel, R, Infostat). Main results: the highest coefficients of molecular similarity were found between: EEB-16 and EEB-21 (0.90), EEB-4 and EEB-10 (0.88), EEB-19 and EEB-25 (0.75), and EEB-2 and EEB-12 (0.73); the 6 highest yield Criollo-type clones were followed up; the morphological consensus was 83.6%, and the final consensus was 94.3%; both, in the molecular configuration and in the final consensus, EEB-23 did not resemble Criollo type clones, being closer to the control representative of the Forastero genetic group and to CCN-51; the other Criollo-type clones formed a separate group from the Forastero, and in general they resembled more the controls of the Nacional and Trinitario genetic groups. The technique was appropriate to process the type of data obtained, so it is recommended to use it in similar studies or in others where it is necessary to analyze different types of variables on the same individuals, simultaneously.