Vegetation and landform relationships at Ajusco volcano Mexico, using a geographic information system (GIS).

摘要

This paper analyses the relationship between distribution of plant communities and landform properties, in an integrated (phyto-geomorphology) form, to define the best sites for reforestation on this highly stressed volcano (Ajusco volcano) located near Mexico City, Mexico. Plant-landform units were determined through the interpretation of aerial photographs (1:20 000 scale), supported by fieldwork and GIS overlay procedures. Vegetation composition and structure analysis were performed using field data collected at 69 plots. A total of 89 species and three environmental variables (altitude, slope and aspect) were analysed using multivariate classification and ordination analysis (TWINSPAN and DCA). A geographic information system was utilized for map digitization, for the generation of a digital elevation model (DEM) and for linking results of multivariate analysis to digital thematic information maps. Multivariate analysis identified eight community types defined by aspect and altitude conditions. The south and north slopes of this volcano showed important differences in vegetation, landforms and rock composition due to past processes (glacial at the north and periglacial at the south) and present conditions. Pine (Pinus hartwegii, Pinus montezumae)-grassland (Festuca amplissima, Muhlenbergia macroura, Calamagrostis tolucensis, Senecio angulifolius, S. barba-jaohannis, Solanum cermtessi) communities were more dominant on south and southeast-facing slopes, while fir (Abies religiosa) communities were more dominant on north-facing slopes, particularly at lower elevations. A combination of spatial and multivariate analysis to study landform and vegetation was very useful to determine the best sites (convex slopes for pine and concave for fir species) for reforestation programmes. An important advantage of this approach is that it allows relatively fast updating of vegetation maps with an appropriate balance between data accuracy and costs..