Abstract

TEICH, I; MOTTURA, M; VERGA, A  and  BALZARINI, M. Association between genetic and phenotypic variability with adjustment for spatial autocorrelation in Prosopis. BAG, J. basic appl. genet. [online]. 2015, vol.26, n.2, pp.63-74. ISSN 1852-6233.

Association studies that relate phenotypic and genetic information allow the detection of relevant interactions. In these studies it is necessary to consider the effects of spatial structure in the data, which is very common in nature, particularly in plant species. Prosopis chilensis and P. flexuosa are two hardwood arboreal species with environmental and economic importance. Significant gene flow occurs between them with the production of natural and fertile hybrids. Given that morphological characters are used to classify individual trees, it is of interest to explore the association between the phenotypic variables used in the taxonomic classification and genetic variability. In this work, multivariate associations between genetic and morphological variability were explored in a P. chilensis and P. flexuosa hybrid swarm considering the possible spatial autocorrelation. For this purpose the molecular (six polymorphic microsatellite markers) and the morphological (13 traits) characterizations of 81 adult trees in a 4,700 m2 plot in Córdoba, Argentina were used. The spatial structures of genetic and phenotypic variability were analyzed at a fine spatial scale. Both characterizations were associated through a Mantel approach based on the correlation of distance matrixes, as well as with new approaches that include the complementary use of multivariate analysis and mixed linear models. A significant genotype-phenotype association, relevant for the management of Prosopis genetic resources was detected. Even with a weak spatial structure, the inclusion of the spatial autocorrelation in the analyses improved the detection and modeling of the association between the observed genetic and phenotypic variability.

Keywords : Spatial Statistics; Moran Index; Partial Mantel Test; Canonical Correlations; Generalized Procrustean Analysis.