Original paper

Zur oberen Waldgrenze in tropischen Gebirgen

Miehe, Gregor; Miehe, Sabine

Phytocoenologia Band 24 Heft 1-4 (1994), p. 53 - 110

108 references

published: Apr 8, 1994

DOI: 10.1127/phyto/24/1994/53

BibTeX file

ArtNo. ESP024002400002, Price: 29.00 €

Download preview PDF Buy as PDF


A review of the literature on the upper forest limit in tropical mountains, results of recent research in the Andes and own fieldwork in southern Ethiopia led to a revision of the current concepts concerning forest and tree limits in the tropics: 1) The upper limit of tropical subalpine forests may well reach altitudes between 4000 and 4300 m a.s.l. 2) Clear-cut treelines, which were regarded as typical of tropical mountains, are largely created by fire on smooth, homogenous slopes. If man-made fires are absent, the forest belt ascends close to the upper limit of phanerophytes: at least in the Bale Mountains/S Ethiopia, the upper forest limit largely coincides with the upper limit of continuous ericaceous vegetation, i.e. the "Ericaceous Belt" sensu Hedberg (1951), as far as it can be concluded from relict quasi-natural vegetation. Depending on the microrelief and the homogeneity of the substratum, the uppermost, 2 to 5 m dwarf forests constituted of single- to multi-stemmed Erica trimera individuals are separated from the adjacent afroalpine dwarf-scrub by a more or less wide mantle of shrubby specimens of the same species. This gradual formation change at the upper limit of phanerophytes on Ethiopian mountains is similar to those observed in temperate forests of the southern hemisphere. One important difference of the temperate forest limits is the snow deformation of the uppermost trees and shrubs. On the tropical mountains, on the other hand, the unique life form of giant rosette plants occurs in the vicinity of the forest limit or above. 3) There is still no evidence of any specific run of the upper forest limit on tropical mountains. In all climatic regions, level valley floors may be avoided by upper montane forests in case of temporary or permanent water stagnation or cold-air-ponding, whereas the higher flanks are still forested. The opposite case occurs in ravines and more inclined V-shaped valleys dissecting wind-exposed slopes: here the forests may climb higher in response to the combined effects of wind- and fire-shelter and more favourable hygric conditions. At least in the Bale Mountains, wind-shelter in such locations seems to be more important for the temperature balance than for the hygric one. 4) Isolated tree groups or groves above actual upper limits of continuous forests are, according to the available field observations, relicts of formerly higher-reaching forests that survived in favourable habitat where they are less vulnerable to fire, grazing and woodcutting. These habitats are not exclusively boulder slopes. The potential, zonal upper limit of forest formations should, however, not be drawn at the uppermost, extrazonal occurrences of trees or groves without hesitation: these highest specimens might owe their existence to microclimatic favour (warmth, humidity), being naturally absent from zonal habitat at comparable altitudes. In the wind-influenced Bale highlands, the altitudinal differences between the zonal and extrazonal upper limits of forests may attain 100, perhaps even 200 m. 5) There are no microclimatic records from tropical mountains which prove the thermic favour of boulder slopes. The only measurements available from the European Alps indicate that the soils of these habitats may be even cooler than those on slopes devoid of rocks. 6) It is not yet proven that there is any dependency of the upper forest limit on the soil temperature on tropical mountains. The available microclimatic measurements indicate that other habitat factors might at least superimpose the possibly limiting effect of low soil temperatures. In the Bale Mts., lack of warmth in the air 0.3 to 3 m above the ground seems to be more important: exposure to wind as a cooling agent causes a depression of upper forest limits. In the whole, there is increasing evidence from different continents that the tropical mountains have been widely transformed into cultural landscapes by man, to a similar extent as it was done by pre-industrial cultures in Europe and High Asia. With the clearing tool of fire stone-age hunters, herdsmen and farmers have comparable influence on high montane forests: taking advantage of the reduced competitive and regeneration power of woody phanerophytes close to their upper limit, man succeeds everywhere in depressing the treeline in order to gain suitable open hunting or grazing ground or arable land. Thus, Ellenberg's (1958, 1979) hypotheses concerning the potential extent of forest in the Andes are supported by a number of recent observations and seem to become valid for all tropical mountains.


tropical subalpine forestssubstratumzonal habitatcultural landscapes