全球变化知识资源中心

Aims: To identify shrubland types of the Bolivian Altiplano based on their floristic composition and on ecological factors. Location: Central and southern Bolivian Altiplano (Bolivia, central-western South America). Methods: Vascular plants were recorded in a field survey of 101 releves (10 m(2)). Releves were subjected to hierarchical agglomerative classification to define numerical vegetation groups. Classification techniques were based on the beta-flexible linkage method (beta = -0.25) with Sorensen distance. The highest crispness values defined the level of the main number of clusters identified. Diagnostic species were identified by means of the phi coefficient of fidelity. Canonical Correspondence Analysis, the Kruskal-Wallis test and a Z test were performed to assess the key ecological drivers of diversity in the Altiplano shrubland vegetation. Results: Based on numerical analyses of phytosociological releves, our work proposes four vegetation types of shrublands in the dry central and southern Bolivian Altiplano. They correspond to the following: tolillares - thickets of Fabiana densa - of the central-southern Altiplano with Junellia seriphioides; tolillares of the central-northern Altiplano with Lobivia pentlandii; lampayares - thickets of Lampayo castellani - with Parastrephia quadrangularis; and tolares - thickets of Parastrephia sp. pl . - with Parastrephia lepidophylla and Junellia minima. The bioclimatic variables were the ones best explaining the distribution patterns of the shrubland vegetation in the dry Bolivian Altiplano. Specifically, they separate the tolillares of the central-northern Altiplano with Lobivia pentlandii - at localities with a higher annual precipitation, annual ombrothermic index, and ombrothermic index of the wettest quarter - from the tolillares of the central-southern Altiplano with Junellia seriphioides. These bioclimatic gradients also position lampayares at localities with a lower than average value of annual precipitation, annual ombrothermic index, and ombrothermic index of the wettest quarter. Significant differences were found when comparing the topographic position, the degree of soil drainage and the frequency of flooding between the vegetation of tolillares on the one hand, and the vegetation of lampayares and tolares, on the other. Lampayares were exclusively related to sandy soils. Conclusions: Our four groups characterize variation within the habitat and elucidate bioclimatic gradients and soil features with related habitats. This knowledge could provide basic information on the vulnerability of different Altiplano shrubland habitats to climatic fluctuations, as this area is highly vulnerable to extreme periods of drought associated with the regional effects of climate change as well as to anthropogenic factors.