Quantitative genetics of wood quality traits in Scots pine

Abstract: Wood quality of commercial tree species is important for many wood processingindustries and thus should be considered for inclusion in forest tree improvementprograms. This thesis evaluated the suitability of various proxy methods for rapid andnon-destructive assessment of wood quality traits on standing trees of Scots pine and thepotential for genetic improvement of different wood quality traits through recurrentselective breeding.Penetrometer Pilodyn and micro-drill Resistograph were tested for non-destructiveassessment of wood density (DENPIL and DENRES, respectively), using SilviScan density(DENSILV) as a benchmark. A strong additive genetic correlation was observed betweenDENSILV and DENRES (rA = 0.96), whilst the correlation with DENPIL was substantiallylower (rA = 0.74). Furthermore, SilviScan stiffness (MOESILV) was used as a benchmarkfor evaluation of several approaches of calculating the dynamic modulus of elasticity(MOE) from standing-tree acoustic velocity (VELTREE). The combination of VELTREEand adjusted DENRES provided the most accurate estimate of MOETREE (rA = 0.91).Additionally, non-destructive acoustic sensing tools were tested at different stages ofwood processing (on standing trees, felled logs and sawn boards) using destructivelymeasured sawn-board stiffness (static modulus of elasticity, MOES) and strength(modulus of rupture, MOR) as benchmarks. They proved to be capable of accuratelypredicting MOES (rA ≈ 0.8) while VELTREE, adjusted DENRES and MOETREE wellreflected MOR (rA ≈ 0.9). Genetic variation of shape stability of sawn boards (bow, crookand twist) was also investigated. Under-bark grain angle (GRA) was found to be a goodpredictor of sawn-board twisting and crooking (rA = 0.84 and 0.62, respectively). Thechemical composition of juvenile wood (proportion of cellulose, hemicelluloses, ligninand extractives) was predicted from Fourier transform infrared (FTIR) spectra usingpartial least squares regression (PLSR) modeling. Individual-tree narrow-senseheritabilities (ℎi2) for all of the studied wood quality traits varied from low to moderate.Genetic improvement of sawn-board DEN, MOES and MOR as the target traits couldbe achieved through selective breeding for MOETREE, DENRES, stem straightness (STR)or GRA. Selection focusing on GRA would also result in lower bow, crook and twist.Despite the negative genetic correlations between growth and wood quality traits, apossibility of their simultaneous improvement was identified. An index combining stemdiameter (DBH) and MOETREE provided the best compromise.