c7.7.36 – Analysis of covariation pattern in leaf morphological variability of several sympatric European white oak species (subgen. Quercus)

Viscosi V (1), Lepais O (2), Gerber S (2), Fortini P (1)

Museo Erbario del Molise – Dip. S.T.A.T. – Università degli Studi del Molise – C.da Fonte Lappone, 86090 Pesche (IS); INRA, Cestas, Université de Bordeaux 1, Cestas, France
Collocazione: c7.7.36 – Tipo Comunicazione: Poster
7° Congresso SISEF *
Sessione 7: “Poster” *

Contatto: Vincenzo Viscosi (vincenzo.viscosi@unimol.it)

Abstract: White oak species (subgen. Quercus) are characterized by high intra- and inter-specific variability in leaf morphology, principally due to the phenomena of hybridization and introgression. In particular, when more oak species live in sympatry, hybrids of different generations show a pattern of morphological variability, which typically is expressed along a morphological continnum between the parental species. In this study, 2-Block Partial Least-Squares analysis was used to study the pattern of covariance among different sets of leaf morphological data, sampled over 336 trees in a mixed forest of Q. petraea, Q. pubescens, Q. pyrenaica and Q. robur, in south-western France: oaks were assigned to pure species or hybrids by genetic assignment. 2B-PLS analysis is a statistical method recently developed to explore patterns of covariance between the variables of two or more data sets and allows constructing linear combinations, which can explain most of the covariance between the measured variables. This work aims to study the patterns of covariance among three morphological data sets, related to leaf shape (landmarks and outlines) and 19 morphological variables traditionally used to study leaf morphology white oak species. For each data set, all variables were reduced to PCs and subjected to 2B-PLS: leaf variability expalined by outlines and traditional morphological variables was analyzed in relation to landmark data. This allowed visualizing, through deformation grids, the shape variation along the more significant linear combinations. In detail, seven significant linear combinations were extracted between outline and landmark data, and between landmarks and the 19 variables. The first two linear combinations distinguished oak species in both analyses: 1) Landmarks vs Outlines – first linear combination explained the differences in length of petiole and in apical shape, the second linear combination was relate to differences in depth of the lobes. 2) Landmark vs 19 variables – the first linear combination was related to the differences in depth of lobes, while the second to the differences in hairiness and in length of the petiole.

Citazione: Viscosi V , Lepais O , Gerber S , Fortini P (2009). Analysis of covariation pattern in leaf morphological variability of several sympatric European white oak species (subgen. Quercus) . 7° Congresso Nazionale SISEF, Isernia – Pesche (IS), 29 Set – 03 Ott 2009, Contributo no. #c7.7.36