Characterizing variations in growth characteristics between Douglas-fir with different genetic gain levels using airborne laser scanning

Abstract

In British Columbia, Canada, Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco] selective breeding is used to develop genetically improved regeneration stock. To evaluate realized performance of improved stock, breeders take frequent measurements in genetic gain trials to determine whether yield gains are being met. Currently, variables collected focus on individual tree yield attributes; however, information generated from progeny test trials may not directly reflect plantation performance. Realized yield trials help to bridge the gap between progeny test estimated gain and plantation setting performance. High density airborne laser scanning (ALS) has the potential to identify variables that could improve the selection and validation process. We utilized ALS collected from an unmanned aerial system to assess performance of genetic improvement across three genetic levels and three stand spacings. ALS derived metrics were used to test three hypotheses: (1) tree height is correlated with the level of genetic superiority, (2) tree crown shape varies across genetic level, and (3) tree crown density is associated with genetic level. Random forest algorithms were used to identify candidate ALS metrics. To account for interaction effects, a two-way analysis of variance was conducted for each metric, followed by a post-hoc test to investigate significant differences between genetic level and spacing. The scale and shape parameters of Weibull probability density functions, vertical complexity index, and the fraction of euphotic voxels were found to be important metrics. Results show that genetically superior trees are typically taller, with higher, shorter and denser crowns. In addition, variation across genetic level may be indicative of greater phenotypic plasticity, as superior trees possess the ability to respond to tighter stand spacing.