Link zu WSL Hauptseite Swiss Federal Institute for Forest, Snow and Landscape Research WSL

Long-term treeline research at Stillberg

Background and research questions

A history of many centuries of seasonal livestock grazing of alpine pastures has significantly depressed the treeline in many European mountain ranges. Therefore, observations of treeline patterns are more likely to be obscured by historical land use than in regions with less human influence and might not provide relevant information about environmental factors that influence mortality and growth patterns at the natural high-elevation limit of tree existence. To overcome this, we used a large afforestation experiment in the Swiss Alps, "Stillberg", to address the following questions:

  • Can trees survive above the current treeline in the European Alps?
  • Are environmental factors that influence spatial patterns of mortality the same as those that influence height growth?
  • Does the relative importance of these environmental variables change over the first 30 years after planting as seedlings?
Versuchsfläche Stillberg    

Stillberg research site photo: Nick Dawes



The 5 ha Stillberg long-term afforestation research area is located near Davos (Central Alps, Switzerland). The bottom of the afforestation is approximately at the current treeline and it covers an altitudinal gradient of more than 150 m (2075 to 2230 m a.s.l.; see picture above). Approximately 92000 seedlings of three treeline species (Larix decidua, Pinus mugo ssp. uncinata and Pinus cembra) were systematically planted in 1975, and mortality and height growth were closely monitored during the following 30 years.

We used decision-tree models and generalized additive models to identify patterns in mortality and growth along gradients in elevation, snow duration, wind speed and solar radiation and to quantify interactions between the different variables.


For all three species, mortality rates were highest during the first 15 years of the study period (Fig. 1). The 30-year period analyzed showed rather high mortality in both pine species (Fig. 1). P. cembra and P. mugo were mainly killed by the pathogenic snow fungi Phacidium infestans (P. cembra only) and Gremmeniella abietina. The years when most trees are completely covered by snow in the early spring, and are therefore most susceptible to attacks by pathogenic fungi (in the case of pine species), may be the most critical stage determining the distribution of trees at the alpine treeline in the Alps at moist sites. Although initial establishment after germination is considered a critical bottleneck for survival, the life stages investigated here can also strongly influence mortality patterns.


Fig. 1. Survival curves for the three species for the period 1975-2005.

The three tree species showed a sharp increase in mortality at elevations above ca. 2160 m a.s.l. (Fig. 2). Height growth of all three species gradually decreased as elevation increased.

Mortalität in Höhenlagen    

Fig. 2. Estimated GAM functions for elevation as a predictor of mortality during three overlapping periods. Each plot shows the relationship of the fitted function with the response, scaled to zero. The plots include approximate 95% pointwise CI bands (in light gray) and red dotted lines mark the elevation of 2160 m where a sharp decrease in survival is observed.


The three species showed a general pattern of lower mortality at locations with an earlier snowmelt date in the spring (highest survival before day of year 140; 20th May). All three species showed a height growth optimum on a later snowmelt date than mortality (ca. day 145-150) (Fig.3).

Mortalität und Schneeschmelze

Fig. 3. Estimated GAM functions for snowmeltdate as a predictor of mortality during three overlapping periods. Each plot shows the relationship of the fitted function with the response, scaled to zero. The plots include approximate 95% pointwise CI bands (in light gray), red dotted lines mark the optimum snowmeltdate for survival and blue dotted lines mark the optimum snowmeltdate for height growth.

Snowmelt date remained the most important variable for mortality and elevation the most important variable for growth over the entire 30 years. Even with the observed significant increase in air temperature during the 30-year period analyzed, elevation remained the most important factor for height growth and the second most important factor for mortality. This result indicates that the increase in temperature over time was not large enough to influence spatial patterns of tree survival and growth. Alternatively, elevation may be associated with other environmental factors in addition to temperature. The increasing importance of wind on tree survival and height growth could be caused by several factors such as an increase in mechanical damage as trees grow taller or by ice crystal abrasion of blowing snow.


  • Established seedlings are capable of surviving more than 100 m above the current treeline in the Alps. However, the observed elevational thresholds in both tree growth and survival indicate a limited potential for the natural treeline to shift upward under the current climatic conditions.
  • Small changes in the duration of snow cover in spring, in addition to elevation, can have a great impact on species mortality and growth patterns at the alpine treeline.


Ignacio Barbeito (SLF)

Melissa Dawes (SLF)

Christian Rixen (SLF)

Peter Bebi (SLF)

Josef Senn (WSL)

More information of this study at:

Keywords Baumgrenze, Wachstum, Überleben, Arve, Föhre, Lärche