Modeling successional patterns of high-elevation forests under changing herbivore pressure – responses at the landscape level

We describe an ecological risk assessment for determining the potential long-term impacts of browsing by red deer, roe deer and chamois on the stand dynamics of mountain forests in xeric parts of Central Europe. The assessment is performed by applying a forest succession model of the FORET/JABOWA type which was supplemented with an empirically based browsing subroutine to simulate the effects of browsing by red and roe deer, and chamois. The following scenarios are assumed: (1) no browsing; (2) actual browsing intensity as determined in field trials; (3) browsing intensity twice as high as in scenario (2); and (4) browsing intensity three times as high as in scenario (2). The simulations for two different forest sites in the subalpine region of Eastern Switzerland revealed that heavy browsing can alter development and composition of the observed forests. The extent of this alteration was influenced not only by high browsing intensity, but also by the selection of tree species which were browsed on a particular site at the same time. Taking into account all the uncertainties of the model approach, we conclude from our preliminary studies that, in the regions considered, even strong herbivore pressure (2–3 times today's values) is not a major threat for the survival of the forest as a biological resource, i.e. the long-term total woody biomass is not altered significantly and the changes in the successional pathways are not dramatic. High browsing intensities, however, alter the forest structure considerably. We found that heavily browsed forests tend to be more open, mortality is higher and trees reach an intermediate and adult size quicker. This shortened development cycle of the stands may have implications for forest functions. The risk assessment yields critical values for mean browsing intensity, ‘MBI', (Pinus cembra/Pinus mugo: 30%; Picea abies: 50%; and Larix decidua: 40%), below which successional patterns and forest structures with browsing are not significantly different from forest succession without browsing.