We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
How to predict tree death from inventory data - lessons from a systematic assessment of European tree mortality models.
- Authors
Hülsmann, Lisa; Bugmann, Harald; Brang, Peter
- Abstract
The future development of forest ecosystems depends critically on tree mortality. However, the suitability of empirical mortality algorithms for extrapolation in space or time remains untested. We systematically analyzed the performance of 46 inventory-based mortality models available from the literature using nearly 80 000 independent records from 54 strict forest reserves in Germany and Switzerland covering 11 species. Mortality rates were predicted with higher accuracy if covariates for tree growth and (or) competition at the individual level were included and if models were applied within the same ecological zone. In contrast, classification of dead vs. living trees was only improved by growth variables. Management intensity in the calibration stands, as well as the census interval and size of the calibration datasets, did not influence model performance. Consequently, future approaches should make use of tree growth and competition at the level of individual trees. Mortality algorithms for applications over a restricted spatial extent and under current climate should be calibrated based on datasets from the same region, even if they are small. To obtain models with wide applicability and enhanced climatic sensitivity, the spatial variability of mortality should be addressed explicitly by considering environmental influences using data of high temporal resolution covering large ecological gradients. Finally, such models need to be validated and documented thoroughly.
- Subjects
GERMANY; TREE mortality; FOREST surveys; FOREST reserves; META-analysis; TREE growth
- Publication
Canadian Journal of Forest Research, 2017, Vol 47, Issue 7, p890
- ISSN
0045-5067
- Publication type
Article
- DOI
10.1139/cjfr-2016-0224