Modelling effective thermal climate for mountain forests in the Bavarian Alps: Which is the best model?

Authors

  • Birgit Reger,

    1. Faculty of Forest Science and Forestry, University of Applied Sciences Weihenstephan-Triesdorf, Hans-Carl-von-Carlowitz-Platz 3, 85354 Freising, Germany
    Search for more papers by this author
  • Christian Kölling,

    1. Bavarian State Institute of Forestry, Hans-Carl-von-Carlowitz-Platz 1, 85354 Freising, Germany
    Search for more papers by this author
  • Jörg Ewald

    1. Faculty of Forest Science and Forestry, University of Applied Sciences Weihenstephan-Triesdorf, Hans-Carl-von-Carlowitz-Platz 3, 85354 Freising, Germany
    Search for more papers by this author

  • Co-ordinating Editor: Robert Peet

  • Reger, B. (corresponding author, birgit.reger@hswt.de) & Ewald, J. (joerg.ewald@hswt.de): Faculty of Forest Science and Forestry, University of Applied Sciences Weihenstephan-Triesdorf, Hans-Carl-von-Carlowitz-Platz 3, 85354 Freising, Germany
    Kölling, C. (christian.koelling@lwf.bayern.de): Bavarian State Institute of Forestry, Hans-Carl-von-Carlowitz-Platz 1, 85354 Freising, Germany.

Abstract

Question: Which thermal climate model performs best in predicting the combined effects of temperature and radiation on forest vegetation in the Bavarian Alps?

Location: Bavarian Alps, Germany.

Methods: In order to find the best model for effective thermal climate for the Bavarian Alps, we analysed models using the following predictors derived from climate data and/or a digital elevation model: (a) temperature variables only, (b) temperature plus slope aspect and inclination, and (c) temperature plus potential global solar radiation. Models were tested by linear regression against four response variables based on average Ellenberg indicator values for temperature (cover weighted/unweighted, with/without bryophytes), which were computed for 2280 georeferenced relevés from the vegetation database BERGWALD. We optimized (b) by empirically searching for thermally most favourable slope aspect and inclination.

Results: Closest model fit was achieved for unweighted temperature values based on vascular plants without bryophytes. Model fit (adj. R2) increased from using temperature alone to temperature–radiation, to temperature–aspect–inclination as predictors. The best spatially explicit model for predicting temperature values (adj. R2=0.57) was based on the variable combination mean temperature in the growing season (May to September), slope aspect (optimal aspect 195°) and inclination (optimal slope 30°).

Conclusion: Combining mean temperatures and relief variables in GIS allows creation of predictive maps of mountain forest response to thermal climate. Applied to climate change scenarios, our model can forecast potential vegetation distribution in the future. The superiority of simple empirical relief factors over a widely used model of potential radiation casts doubt on the meaningfulness of the latter for vegetation studies.

Ancillary