Genetic Variability of Fifteen Romanian Provenances of Norway Spruce (Picea abies (L.) Karst.). Preliminary Results

1991

S. 179-192

5 Abb., 9 Tab., 13 Lit. Ang.

Unselbständiges Werk

200041311

Annales des Sciences Forestieres 1991 ; 48,Nr. 2, S. 179-192

A collection of 15 Romanian provenances of Norway spruce has been studied: 13 are natural and originate from Bucovine (north-eastern), Bihor (north- western) and central Romania; 2 others are artificial. They are represented by 118 open-pollinated progenies and compared with 3 standards: 2 French provenances and a polycross. The material was sown in 1979 and planted in 1982 at 2 sites located in the Eastern part of France (Gennes, near Besancon; and Amance, near Nancy). The traits studied here are the height at 10 yr, bud burst in 1986, branch angle and forkness. Compared to the standards, the Romanian provenances are fast-growing, relatively late flushing (except the 2 artificial ones) and have a good branch angle. The provenance effect for this last trait seems to be weaker than for growth or flushing (table III). Forkness (table V) has an unfavourable correlation with total height (r = 0.55) and branch angle (r = 0.57). The provenance x site interaction for height is significant, but cuased by one unstable provenance only (fig. 3). Two stable provenances may be selected: 1 from Bucovine and 1 from Bihor. Compared to the best standard, they give a gain of 27% for total height, 13% for branch angle, and with almost the same flushing time as the latest standard (table VI). The study of genetic parameters shows a lack of variability between families for 2 provenances originating from the central area. According to the variances-covariances homogeneity, the other provenances are pooled into 2 groups in Gennes (one is composed of the 2 artificial provenances) and only 1 group in Amance. Narrow sense heritability (table VII) is 0.16-0.37 for branch angle, 0.32- 0.46 for height, and is a litter higher for flushing (0.24-0.78). Genetic additive correlations between traits are generally favourable but not significant (table VIII). Forkness has an unfavourable correlation with total height (0.25) and branch angle (0.36), as at the provenance level. The family x site interaction is only significant for height, but is caused by only one unstable family. The expected genetic gains for a clonal seed orchard option are estimated by a multi-trait combined individual and family selection and compared to the best standards, the 50 higher indix give an expected genetic gain of 34% for 10 yr height and 17% for branch angle and are as late as the Chapois standard (table IX). These results show the interest of some Romanian provenances particularly those from Bucovine and Bihor, but this should be confirmed at a less juvenile stage and for other traits.