The Ecology & Biodiversity Group, Utrecht University, 3512 JE Utrecht, THE NETHERLANDS.
This thesis reports on the first systematic sampling, species identification and analysis of community composition of epiphytic bryophytes in nine localities across the Amazon basin. Eight trees were sampled per locality, from the base to the outer canopy in five height zones. The sampling resulted in 3104 records, 3066 of which resulted in the identification of 225 species and 38 morpho-species and 40 records remained unidentified. The species belonged to 29 families. The most common families in number of records were Lejeuneaceae (55%), Calymperaceae (8%), Leucobryaceae (4%) and Sematophyllaceae (4%). Over the entire Amazon bryophytes showed a clear compositional gradient along the host trees according to the height zones, supporting strong niche assembly of communities at local scale. When the community composition of the entire localities were analysed, however, species richness and relative species abundances of epiphytic bryophytes in most of the nine Amazonian localities studied were consistent with the Neutral Model of Biodiversity and Biogeography (Hubbell 2001). Probably this was due to the lack of a strong compositional gradient pertinent to epiphytic bryophyte vegetation at a large spatial scale. Therefore, the composition of a given locality is mainly regulated by stochastic recruitment of individuals from the same locality or from immigration, driven by the abundances of the species they belong to in the metacommunity. The exception to the neutrality across the Amazon is given by the locality in Ecuador, which has a significantly higher number of species. Based on simulations, we predicted that the average epiphytic bryophyte community of 250 occurrences found in a sample of 8 trees in the Amazon is expected to have around 65 species including 21 singletons, with the most common species having between 10-30 occurrences. The general picture given by the results is the combination of strong niche assembly in different height zones on a host tree to the composition of an entire locality driven by neutral dynamics. Ultimately, the composition of a community of epiphytic bryophytes in a given height zone of a host tree in a given locality is driven by the following pattern: The environment has a stronger role at the extremes of the gradient, so that communities of zones 1 and 6 – bottom and canopy – have the strongest influence of establishment limitation. The influence of dispersal overrides establishment limitation in the middle trunk zones 2, 3 and 4. In all zones, the possibility of a species to occur is significantly higher for species already established in the surroundings of the same locality than for species arriving through long distance dispersal. Finally, the chance of a species not yet present at the locality of the community to occur, i.e. through long distance dispersal, significantly increases with increasing height zone of the community. Furthermore, two predictions were tested in the most abundant species of the dataset, Cheilolejeunea rigidula, using molecular data as a measurement of distance among individuals. The results were in agreement with the predictions of predominance of local recruitment and lack of geographical structure in the immigration across the basin.