On the Ecology of Mountainous Forests in a Changing Climate: A ...

Behaviour of FORCLIM along a transect in the European Alps 103
The results obtained from FORCLIM-P at the site Sion do not correspond to phytosociological
expectations (Fig. 4.7): Although this site is very xeric, an extremely high biomass
is attained; while the occurrence of Q. robur is plausible (Ellenberg 1986), the codominance
of chestnut (Castanea sativa) may be questionable, and the considerable biomass
of yew (Taxus baccata) is unrealistic as well.
The litter production simulated by FORCLIM-P is summarized in Tab. 4.4. In coniferous
forests, leaf litter of low quality is produced (uFL 3 ), whereas deciduous forests are characterized
by more easily degradable leaf litter (uFL 2 ). There is some literature data to
evaluate the simulated pattern of total aboveground litterfall: For boreal forests (comparable
to the sites Davos and Bever), Ajtay et al. (1979) give 5.5–6 t·ha -1·yr-1 , while their
value for temperate forests (Bern, Sion, Locarno) is 8.5 t·ha -1·yr-1 ; the rates simulated
by FORCLIM-P are slightly lower. On the other hand, Cox et al. (1978) found only
3.3 t·ha -1·yr-1 in a Liriodendron forest in Tennessee. For coarse woody debris (corresponding
roughly to uWL), Harmon et al. (1986) list a range of 0.17–7 t·ha -1·yr-1 in
coniferous forests, and up to 14.5 t ha -1·yr-1 in deciduous forests. The simulated aboveground
litter production (Tab. 4.4) agrees with the range of data from these sources (cf.
also Vogt et al. 1986).
4.3.2 FORCLIM-E/P
In this model setup, the abiotic environment is stochastic; thus, it is not restricted to
average conditions but includes some of the natural variability of the weather and the
habitat. As in section 4.3.1, available nitrogen is kept constant at 100 kg/ha. The results
of these simulations are shown in Fig. 4.8 & 4.9.
At the site Bever (Fig. 4.8), the spruce forest simulated by FORCLIM-P (Fig. 4.6) is replaced
by the Larici-Pinetum cembrae, where P. cembra dominates (Ellenberg & Klötzli
1972). Picea excelsa is not competitive any more due to the occurrence of summer
droughts. On northern slopes at Bever (Fig. 4.12), a larch-spruce forest is simulated,
corresponding to the pattern observed in the area (Ellenberg & Klötzli 1972).
The forest simulated at the site Davos (Fig. 4.8) does not differ much from the one simulated
by FORCLIM-P (Fig. 4.6); it still belongs to the Larici-Piceetum (Ellenberg &
Klötzli 1972). Silver fir (Abies alba) occurs now because winter temperature in some
years is high enough to enable establishment of the species. Its presence at elevations