Provenance by site interaction and stability analysis of European beech (Fagus sylvatica L.) provenances grown in common garden experiments

Fifteen provenances of European beech (Fagus sylvatica L.) were evaluated for stability and adaptability by height growth at four test sites in Bosnia and Herzegovina (1), Croatia (1) and Serbia (2). Provenance trials were established in spring 2007 by planting 2- and 3-year old seedlings and arranged in a randomized complete block design. The data from seven 6-year-old and eight 5-year-old provenances were obtained in 2009 and analyzed separately. Finlay and Wilkinson's regression analysis and Additive main effects and multiplicative interaction model (AMMI model) were used to assess provenance by site interaction (P x S) and to identify beech provenances that have high and stable performance in different environments, at the juvenile stage of development. Analysis of variance showed that effects of provenance, site, and P x S interaction are highly significant (p < 0.001) in both age groups. Linear regression model evidenced that most of the provenances had regression coefficients not significantly different from unity (b=1.0), except for provenances Sjeverni Dilj Caglinski (HR24) and Vranica-Bistrica (BA59). The partitioning of the total sum of squares (SS) exhibited that the site effect was the predominant source of variation in both age groups of provenances (50.7% and 38.5%, respectively). Additionally, regression analysis explained 15.8% and 33.2% of provenance by site interaction terms, in provenances age six and five years, respectively, while the AMMI analysis accounted for 62.2% and 78.7% in P x S interaction. The results of AMMI showed that the first principal component (PC1) was statistically significant in both age groups. Adaptability and stability of provenances to the test sites were estimated with AMMI1 and AMMI2 biplots. Provenance Sjeverni Dilj Caglinski (HR24) showed constant performance over tested sites, characterizing with above average height growth at low yielding environments. Provenances HR25 and BA61 showed the opposite type of adaptation, being adapted to high yielding sites. Provenance Valkonya (HU42) was characterized by IPCA1 score close to zero and above average mean height growth, suggesting general adaptation to the tested environments. The implication of P x S was discussed in light of impact of climate change on beech and selection of most suitable provenances for future reforestation programs.