A reconstruction of Colombian biomes derived from modern pollen data along an altitude gradient

• Published in: Review of palaeobotany and palynology 2001-10, Vol.117 (1-3), p.79-92
• Main Authors: Marchant, R., Berrio, J.C., Cleef, A., Duivenvoorden, J., Helmens, K., Hooghiemstra, H., Kuhry, P., Melief, B., Schreve-Brinkman, E., Van Geel, B., Van Reenen, G., Van der Hammen, T.
• Format: Article
• Language: English
• Subjects: Leerstoelgroep Resource Ecology; PE&RC; Resource Ecology
• ISSN: 0034-6667; 1879-0615
• DOI: 10.1016/S0034-6667(01)00078-1

Biomes are reconstructed in Colombia from modern (core-top) pollen data derived from 22 sites along an altitudinal gradient (2000–4100 m) that encompasses the tree line. The `biomization' methodology is described in a stepwise manner that details the reconstruction of vegetation along an altitudinal gradient. In the majority of the cases, the results are comparable to site-specific descriptions of the vegetation. At altitudes between 2000 and 3000 m, cool mixed forest and cool evergreen forest biomes are important. Between 3000 and 3700 m the cool grassland/shrub biome is dominant. Above 3700 m, the affinity score to arboreal biomes is low with a concomitant increased affinity to the cool grassland biome that co-dominates along with the cool grassland/shrub biome. Hence, the technique is shown successful at reconstructing modern vegetation, this is particularly so when affinity scores to the range of biomes, rather the single most dominant biome, are used to describe vegetation composition. The results are able to delimit the altitudinally-induced changes in vegetation, from Andean forest associations, open forest and shrub grassland complex through to high altitude grasslands. However, this ability of the modern pollen data to predict the potential vegetation, via the biomization technique, is complicated by several site-specific factors that impact on the vegetation. Of particular importance are localised environmental conditions, particularly moisture regime. An additional area of interest is the impact that human activity has had on the vegetation. By manipulation of the input matrices, which are assigned a priori, and consultation of the range of biome affinity scores, it is possible to identify which sites reflect an anthropogenic signal and how this is manifested within the pollen data. By investigating the range of affinity scores, these factors can be identified and the relatively subtle changes in reconstructed vegetation determined, in particular, how the composition and quantity of the arboreal component changes about the tree line. The results are discussed in the context of using tree line position, and altitudinally-induced floristic changes, for palaeoclimatic reconstruction.