Multivariate analysis for assessment of the tree populations based on dendrometric data with an example of similarity among Norway spruce subpopulations

Matějka K. (2017): Multivariate analysis for assessment of the tree populations based on dendrometric data with an example of similarity among Norway spruce subpopulations. J. For. Sci., 63: 449–456. The new method for evaluation of tree populations presented here is based on a correlation analysis within a set of dendrometric variables. The correlation analysis is carried out for each population separately. The method evaluates differences between resulting correlation matrices. These distances can be used by hierarchical cluster analysis (unweighted pair-group average) or by ordination analysis (non-metric multidimensional scaling – NMS). Test data were obtained in 10 research plots in the area of Medvědí Mt., Šumava National Park. Plots are located in Norway spruce [Picea abies (Linnaeus) H. Karsten] climax forests. The results enable ecological interpretation of both classification and NMS. The populations (subpopulations) differ in their origin (spontaneous succession or partial planting) and environmental conditions (extreme environment near the mountain summit versus water-logged soils). These differences were reflected in results of the classification and ordination of the spruce (sub)populations.