Defaut de fourchaison chez le pin laricio de Corse (Pinus nigra Arn. ssp. laricio var. Corsicana Loud.) : decomposition du phenomene en vue de l'amelioration genetique

anglais. Two types of forking defects, fork and ramicorne can seriously affect stem quality of Corsican Pine. The use of a forking score between 0 and 100 as selection criterion seemed not to be very efficient for the breeding population studied (tabl. 1). The purpose of this work is to find through factors explaining forking, one or several good predictors ensuring maximum genetic gain. Measurements were made in fall 1985 on a 9 years old open pollinated progeny test located in Orleans forest. The importance of three already known explicatim factors of forking defects was studied for the years 1983 and 1984 on a sample of 443 forked trees. - Insect attacks (Rhyacionia buoliana) and rust damages, - Polycyclism, through the total number of lateral lammas shoots (PA) and lateral lammas shoots inducing defects (PAD), - Average branch angle with a special notation scale (fig. 1).Genetic parameters (family and residual variances, individual heritabilities, genetic and residual correlations) were computed on 48 half-sib families after an analysis of variance with a hierarchical model (tabl. 2). Insect attacks and rust damages can’t explain forking defects either in 1983 or in 1984. Although it is affected by climatic conditions, polycyclism seems to be the main explanatory forks of (presence of forks and ramicornes) in 1983 and 1984. Meanwhile, defects are rather due to a secondary elongation of some lateral buds, than to real lammas shoots (fig. 3). The branch angle score used, discriminates well the most forking trees (fig. 5) and can take into account factors unducing forks independent of polycyclism. This measure is more stable from one year to another and better describes a characteristic of each tree. In the studied breeding population, polycyclism and branch angle explain a great proportion of the observed variability for the forking score (tabl. 4). Meanwhile, because of their high values of between families variances and individual heritabilities, these 2 traits seem to be more interesting for the breeder than the forking score (tabl. 5). It results, from the non significative genetic correlation between total height and branch angle that mark a simultaneous genetic gain can be obtained on both traits (tabl. 6). This progress is limited by the genetic positive unfavorable correlation between polycyclism and total height and the best selection compromise need the use of selection index combining the 3 characters.

français. Une décomposition du défaut de fouchaison chez le Pin laricio de Corse a été réalisée sur les verticilles 1983 et 1984 d’un test de descendances maternelles âgé de 9 ans, situé en forêt d’Orléans. La présence de pousses secondaires ou plus précisément d’allongements secondaires sur les branches latérales, bien que sensible à l’effet année, semble être le principal facteur explicatif de l’apparition des 2 types de défauts : fourches et ramicornes. L’utilisation complémentaire d’une note d’angle moyen de branchaison, plus stable dans le temps, permet d’évaluer l’importance des causes de la fourchaison autres que le polycyclisme. Dans la population d’amélioration étudiée, ces 2 caractères s’avèrent être de bons critères prédicteurs de la fourchaison. avec une variabilité génétique intéressante et une héritabilité individuelle assez élevée (0,3-0,4). L’absence de corrélation génétique significative entre hauteur totale et angle de branchaison réduit l’importance de l’antagonisme vigueur-niveau de fourchaison à la seule corrélation génétique positive entre hauteur totale et polycyclisme. Le meilleur compromis au niveau de l’espérance des gains génétiques sera obtenu grâce à la construction d’un indice de sélection combinant les 3 caractères : hauteur totale, polycyclisme et angle de branchaison.