气候变化领域集成服务门户(CCPortal): How structural diversity affects Norway spruce crown characteristics

CCPortal

DOI	10.1016/j.foreco.2020.117932
	How structural diversity affects Norway spruce crown characteristics
	Bianchi S.; Siipilehto J.; Hynynen J.
发表日期	2020
ISSN	0378-1127
卷号	461
英文摘要	Crown size is an important predictor for forest growth dynamics. Models have been widely used for predicting this information when missing. The static approach simulates the height of the crown base (HCB) or the live crown ratio (LCR) at specific points in time. In Finland, there are species-specific static models for determining LCR in even-aged stands (or rotation forestry, RF), but they are not adequate for uneven-aged forestry (or continuous cover forestry, CCF). We considered a dataset comprising ~32,000 repeated measurements on ~9000 Norway spruce (Picea abies (L.) Karst.) trees, belonging half to RF stands and half to CCF. We simulated HCB and LCR at any measurement time, using non-linear mixed models as function of various tree, stand, site and management variables. We validated them using a leave-one-out cross-validation applied at plot level (111 iterations). We observed differences in the crown length characteristics across the management systems. HCB increased more sharply in RF stands than in CCF stands with tree size and stand development. Generally, tree-level characteristics (tree size and asymmetric competition) were less important than stand-level characteristics (stand development and basal area) in RF stands, and the other way around in CCF. Both models included an indicator of the forest structural diversity as interaction term with all the predictors to account for such differences. After cross-validation, the mean absolute percentage error for the HCB model was 32% and for LCR 15%. The fit was generally better for RF stands than for CCF stands. Our study highlights differences in the crown parameters under different forest managements. The static HCB and LCR models we present here can be used to predict crown parameters in both RF and CCF management systems, although there are limitations in the data range. Even without a real independent validation, we deemed the cross-validation employed robust enough. © 2020 Elsevier B.V.
关键词	Fading (radio)Information management Plants (botany)Statistical methods Timber Asymmetric competition Continuous cover forestries Leave-one-out cross validations Management systems Mean absolute percentage error Picea Abies (L.) Karst Repeated measurements Structural diversity Forestry coniferous tree data set forest management hexachlorobenzene model validation Characteristics Forestry Management Models Picea Abies Statistical Methods Trees Finland Picea abies
语种	英语
来源机构	Forest Ecology and Management
文献类型	期刊论文
条目标识符	http://gcip.llas.ac.cn/handle/2XKMVOVA/132957
推荐引用方式 GB/T 7714	Bianchi S.,Siipilehto J.,Hynynen J.. How structural diversity affects Norway spruce crown characteristics[J]. Forest Ecology and Management,2020,461.
APA	Bianchi S.,Siipilehto J.,&Hynynen J..(2020).How structural diversity affects Norway spruce crown characteristics.,461.
MLA	Bianchi S.,et al."How structural diversity affects Norway spruce crown characteristics".461(2020).