ABSTRACT
The history of Sunflower domestication and especially the selection of high seed oil content explain several
bottlenecks that reduced the genetic diversity of the crop compared to the wild relatives. The underlying
genetic control of a crop's interaction with its environment is complex and often poorly understood. Studies of
geographic and climatic variation can provide important information regarding patterns of adaptation to
different habitats and ecological isolation. The objective of the present study was to determine the patterns of
distribution of morphological variation for thirteen quantitative characters in 77 accessions of wild sunflower
evaluated in the same site.
The collection of genetic resources of wild H. annuus, as held at the French National Institute for Agricultural
Research (INRA) in Montpellier, includes approximately 350 populations, which have been collected in the
USA. Seventy – seven wild sunflower accessions from this collection were compared in France (Mauguio) for
13 quantitative characters using multivariate methods, including clustering, principal component and
canonical discriminate analysis to assess the patterns of morphological and climatic variation. Geographic and
climatic data from their sampled sites were analyzed and we sought for correlations of climatic variables and
morphological traits. Populations from northern Mexico and southern Canada were not available at that time.
Ward's hierarchical clustering (Ward, 1963) was used to identify groups of climatic sites and of morphology
forms. Hierarchical clustering was applied by SAS software (SAS, 2000). Principal Component Analysis
(PCA) (PRINCOMP procedure, SAS software) was performed on the correlation matrix of the morphologic
and climatic variables bases on the mean accession value. Canonical discriminate analysis (CDA) on the basis
of climatic clustering was used to assess the differences between morphological variations of the 77
accessions for the thirteen ch