Stability across sites of Douglas-fir provenances in northern Spain

Abstract

This study examines the stability across sites for height growth at different ages in a multi-site Douglas-fir provenance test in Northern Spain in order to study its implications for provenance use recommendations and breeding strategies. A total of 89 IUFRO provenances, planted on 16 sites, were analysed. Type B correlations were used to examine the relation among geographical differences between sites and their site-to-site correlations. The site index differences between sites was the only variable that explained the Type B correlation variation, indicating that P x E interaction arises from site index differences between test sites. The lack of significant correlations between geographical differences and Type B correlations between sites, suggested that all the studied area should be considered as a unique breeding zone. Despite the relative high Type B correlation within all the studied area (r, > 0.6), the analyses of variance indicated a high relative importance of the P x E interaction, especially at the earlier stages. This interaction must be considered for selection and provenances recommendation. It was concluded that selection should be made for stability and overall good performance within the whole area. Provenance stability over the site index variation was analysed by the joint regression analysis. Significant stability differences were found among provenances. Considering both the overall performance and the stability, 11 provenances were recommended for Northern Spain. Most of these provenances come from North Oregon and South Washington, from latitudes north of 45 " N.