Does the phenotypic selection affect the genetic structure and diversity? A study case on Walnut in eastern central Italy (the region of Marche)

The Persian walnut (Juglans regia L.) is widely planted in western Europe, either for fruit either for high quality timber production. This tree is generally considered non authoctonous, probably introduced from East some 7000 years ago and spread by several ancient civilisations. The possible artificial origin seems confirmed by the low intra-specific variation and the higher individual variability recorded by several Authors as well as by the lack of natural populations. Indeed, only wider fruit cultivation areas or small groups, lines or isolated walnut trees can be recorded in Italy. The occurrence of walnuts in forest, escaped from cultivation areas, is very rare. Due to the increased interest of planters, walnut plantations have been extended several ten thousands hectares throughout all western Europe. As a consequence of that it was evident the necessity of selected suitable basic populations in order to supply high quality reproductive materials. The conventional method based on the organisation of a wide and exhaustive seed procurement from the native range to establish provenance tests is at the present impossible. Thus it is necessary to study methods of selection which consider basic materials growing within the western European range. This study is aimed to test the efficiency of the multi-trait Selection Index method, in preserving levels of genetic diversity and structures compatible with the standards observed within a reference system of extended Italian populations. As a consequence of the relatively recent introduction, the genetic structure of the species shows individual variation higher than inter-population diversity. Those genetic structure characteristics were revealed also during a survey of walnut resources in the region of Marche, central Italy. The survey was the starting point for selecting and preserving basic materials for high quality woody production, possibly interesting for forest nurseries in the region. The genetic variation of Marche’s population, compared to a reference system of 7 other Italian provenances, was used as a base to establish a possible improvement strategy together with basic guidelines to manage those genetic resources. Indeed, the very important individual component of the genetic variation suggested to select directly superior phenotypes in view of establishing a comparative multisite progeny test network. No substantial differences were detected concerning the genetic structure of the Italian population, neither within the Marche population, neither in the phenotypically selected material. Homozygosity was always high, probably due to genetic erosion, isolation and adaptation to extreme conditions. Given this general situation in the Marche area, a special care should be payed in the management of walnut reproductive materials, in order to maintain sufficient levels of variation in plantations.