HAL will be down for maintenance from Friday, June 10 at 4pm through Monday, June 13 at 9am. More information
Skip to Main content Skip to Navigation
Conference papers

Application of Finnish Scots pine branching models on French forest resource: preliminary results

Abstract : In France, the construction of Scots pine growth and yield models at the tree level started in 1999. In order to link growth and yield models with models describing wood quality, additional models describing branch attributes are needed. Branch angle (from vertical) and branch diameter models were already developed for pure Scots pine stands in southern and central Finland ("Finnish" models; Mäkinen and Colin, 1998; Mäkinen et al., 1999). The objective of this study is to answer two questions: - Is it possible to use the Finnish branch diameter and angle models on French Scots pine? - If not, is it possible to limit the number of measured branches necessary to develop new models? To rapidly assess the relevance of using unmodified Finnish models in France, 27 trees were sampled in 3 experimental plantations: a provenance test in Alsace (northeastern France) and 2 thinning experiments in Region Centre. The observed data were compared to the predictions obtained with the Finnish angle and diameter models. The predicted branch angles were always higher than the observed branch angles, the Finnish Scots pines having larger angles than the French Scots pines. The differences in branch angles were interpreted mainly in terms of genetic origin. The predicted branch diameters were underestimated; this underestimation increased with tree size. So, the parameter estimates were not accurate for the stem diameter and the relative crown length for the French data set. The assessment of the Finnish models using the French sample was quite encouraging, especially for the branch diameter model: the mathematical form of the model seemed to be preserved. Conversely, the branch angle model seemed less relevant; a complete development of a new model appeared necessary. To reduce the time and work required for additional sampling, it would be desirable to reduce the number of branches measured on each tree. Toward this goal, the Finnish models were re-estimated using our initial data set, and again successively on reduced data sets in which all branch measurements for x whorls were removed: for example, 1 whorl out of three. The results showed that the loss in accuracy was low if the measured whorls were well distributed along the stem from the apex to the ground, or when the whorls corresponding to specific crown regions were selected. The implementation of this reduced-measurement procedure will allow fewer whorls per tree, and more trees and stands, to be sampled. Our sampling strategy in our new data set will take into account the full range of the regressor variables identified by the Finnish models.
Document type :
Conference papers
Complete list of metadata

https://hal.inrae.fr/hal-02581463
Contributor : Migration Irstea Publications Connect in order to contact the contributor
Submitted on : Thursday, May 14, 2020 - 9:09:02 PM
Last modification on : Thursday, March 31, 2022 - 4:20:31 AM

Identifiers

  • HAL Id : hal-02581463, version 1
  • IRSTEA : PUB00011654

Collections

Citation

Céline Meredieu, Sandrine Perret, Catherine Bastien, F. Colin, Christian Ginisty. Application of Finnish Scots pine branching models on French forest resource: preliminary results. IUFRO workshop S5.01-04 on Connection between silviculture and wood quality through modelling approaches and simulation software, Harrison, CAN, 8-15 September 2002, 2002, Harrison, Canada. pp.391-398. ⟨hal-02581463⟩

Share

Metrics

Record views

9