Gottingen Centre for Biodiversity and Ecology, University of Gottingen, 37073 Gottingen, GERMANY.
This dissertation focuses on two aspects of tropical forests. In the first part I report patterns of plant diversity at local scales in a central Bolivian Andean foothill seasonal forest and evaluate the importance that different life-forms have on contributing to overall diversity. Total vascular plant species surveys from three 1-ha plots yielded species richness values from 297 species and 22360 individuals per hectare to 382 species and 31670 individuals per hectare. Epiphytes, and other non-woody life-forms, contributed to significant numbers of overall species richness and abundances. Comparing the observed patterns with other inventories in the Neotropics, showed that the studied Central Bolivian forest plots were similar in total species richness to other dry deciduous and humid montane forests, but less rich than most Amazonian forests. Nonetheless, species diversity of lianas, terrestrial herbs and especially epiphytes proved to be of equal or higher species richness than most other neotropical forest inventories from which data are available. These results highlight the significant contribution that non-woody life-forms in Andean forest ecosystems have toward overall species diversity and abundances, and show that we need an increased inventory effort of these life-forms in order to obtain accurate information useful for the characterization of vegetation types, for the mapping of diversity hotspots and ultimately for conservation purposes.
However, since woody species (especially trees above a certain diameter cut-off) are still extensively used to characterise tropical vegetation, I also present a detailed account of the tree inventories on the same three one-hectare plots in central Bolivia. Inter-plot comparisons showed remarkable variation even though the vegetation of the study area was assigned by a recent classification to one single vegetation unit. Few species were shared among plots and most (between 34% and 50%) were locally rare, i.e., species with only one or two individuals per plot. The species richness values we found in this study were similar to other tree inventories in comparable seasonal forest ecosystems in Bolivia. Species and familial composition, however, were contrastingly different, except for the well-known fact that Leguminosae is the numerically most important family in seasonally dry neotropical forest ecosystems.
In the second part of this dissertation I focus on (meta-)community patterns and address questions about the processes and mechanisms that might have produced them. I first present results from a study at a large geographical scale, covering the whole of Bolivia. Using presence-absence data for species of Acanthaceae, Bromeliaceae, Cactaceae and Pteridophyta occurring in Bolivian Andean seasonally dry forest islands I explore patterns in the beta-diversity of these plant groups. Floristic comparisons among the islands showed that Acanthaceae and Bromeliaceae, showed coincident biogeographic patterns, suggesting two disjunct seasonally dry forest groups in Andean Bolivia: one including all small isolated northern dry valleys and another including all southern valleys with connections to the lowland seasonal forests in southern and western Bolivia. Furthermore, the analysis of the variation of the beta-diversity of each studied plant group suggested an important role of group-specific dispersal characteristic. Thus, plant groups with species that have seed dispersal restricted to short distances (a few tens of metres as in Acanthaceae and Bromeliaceae) were geographically structured. In contrast, groups with species without long-distance dispersal limitation and with a potentially ubiquitous distribution (as in pteridophyta, due to their wind-dispersal system), were rather more influenced by local environmental site conditions suggesting post-dispersal restricting mechanisms (e.g., during establishment).
I a second study, I used the information from the total vascular plant inventories in central Bolivian forest referred to above to evaluate if and to what extent a suite of environmental factors influenced plant species richness and community composition at the local scale. The results suggested a major role of selected above and below-ground environmental gradients in determining small scale patterns of species richness and community composition, of the whole forest community and also of each life-form group (terrestrial herbs, epiphytes, shrubs, lianas and trees). Different life-forms, however, were related to different combinations of these factors and the latter were not able to account for a significant fraction of the variation in the data. In conclusion, spatial and environmental factors (individually or acting together) can be invoked to explain species richness and community composition patterns in the tropical deciduous forests of Bolivia. The importance of each one appears to depend on the geographical scale at which the study is made and on the ecological characteristics of the study group. However, neither of them are able to explain all the variation in the data, and novel methods that include evolutionary information of each taxon need to be explored.