Topographic and spatial controls of palm species distributions in a montane rain forest, southern Ecuador

The northern Andes harbour a flora that is as species-rich or even richer than the 18-times larger lowland Amazon basin. Gaining an understanding of how the high species richness of the Andean region is generated and maintained is therefore of particular interest. Environmental sorting due to elevational gradients in climate has been emphasized as a driver of vegetation distribution and plant community assembly in tropical mountain areas such as the Andes for two centuries, while alternative mechanisms have been little studied. Here, we investigated the importance of topography and spatial location as factors controlling species distributions in a palm community in a montane rain forest landscape in the Andes of southern Ecuador (1900-2150 m above sea level). Eleven species were present: Aiphanes verrucosa, Ceroxylon parvifrons, Chamaedorea pinnatifrons, Dictyocaryum lamarckianum, Euterpe precatoria, Geonoma densa, Geonoma orbignyana, Geonoma paradoxa, Prestoea acuminata and Wettinia aequatorialis. To study their spatial distribution, forty 250 m(2) (5 x 50 m(2)) plots were laid out perpendicular to four paths that were categorized into three areas and two topographic units (ridges and gullies). Mantel tests and indicator species analysis showed that both topography and spatial location imposed strong controls on palm species distributions at the study site. Our results suggest that species distributions in the studied montane forest landscape were partly determined by the species' habitat requirements, but also by unknown spatial effects. Although a number of possible explanations exist for the latter, such as unmeasured environmental variables and historical disturbance events, we believe dispersal limitation is likely to be involved. Furthermore, although the gully- or ridge-association of some species corresponded to their general elevational ranges in southern Ecuador, this was not the case for other species. Based on such considerations, we conclude that elevational climatic gradients are likely to only form part of the explanation for the topographic effects on palm species distributions at the study site. Other factors must also be involved, notably wind-exposure and hydrology, as discussed for lowland palm communities. Our results show that to understand plant community assembly in the tropical montane forests of the Andes it is too simple to focus just on environmental sorting by elevational climatic gradients.