Tree mortality in dynamic vegetation models - A key feature for accurately simulating forest properties

Corina Manusch, Harald Bugmann, Caroline Heiri, Annett Wolf

Research output: Contribution to journalArticlepeer-review

39 Scopus citations


Dynamic vegetation models are important tools in ecological research, but not all processes of vegetation dynamics are captured adequately. Tree mortality is often modeled as a function of growth efficiency and maximum age. However, empirical studies have shown for different species that slow-growing trees may become older than fast-growing trees, implying a correlation of mortality with growth rate and size rather than age. We used the ecosystem model LPJ-GUESS to compare the standard age-dependent mortality with two size-dependent mortality approaches. We found that all mortality approaches, when calibrated, yield a realistic pattern of growing stock and Plant Functional Type (PFT) distribution at five study sites in Switzerland. However, only the size-dependent approaches match a third pattern, i.e. the observed negative relationship between growth rate and longevity. As a consequence, trees are simulated to get older at higher than at lower altitudes/latitudes. In contrast, maximum tree ages do not change along these climatic gradients when the standard age-dependent mortality is used. As tree age and size determine forest structure, our more realistic mortality assumptions improved forest biomass estimation, but indicate a potential decline of carbon storage under climate change. We conclude that tree mortality should be modeled as a function of size rather than age.

Original languageEnglish
Pages (from-to)101-111
Number of pages11
JournalEcological Modelling
StatePublished - 24 Sep 2012
Externally publishedYes


  • Climatic gradients
  • Intrinsic mortality
  • Maximum diameters
  • Maximum tree age
  • Vegetation modeling


Dive into the research topics of 'Tree mortality in dynamic vegetation models - A key feature for accurately simulating forest properties'. Together they form a unique fingerprint.

Cite this