Floristic composition in three different forest types in western Amazonia

Aim: The latitudinal gradient is considered a first-order biogeographical pattern for most taxonomic groups. Yet, latitudinal changes in plant ecological communities are not always consistent, and this could be related to the physical and biological characteristics of different forest types. In this study, we compare latitudinal changes in floristic diversity (alpha diversity), composition (beta diversity), and dominance across different tropical forest types: floodplain, terra firme, and submontane forests. 
Location: Western Amazonia (Ecuador, Peru, and Bolivia).
Taxon: Woody plants. 
Methods: We inventoried 1,978 species and 31,203 individuals of vascular plants with a diameter at breast height ≥ 2.5 cm in 118 0.1-ha plots over a 1,800-km latitudinal gradient in three different forest types. The relationships between alpha diversity, latitude, and forest type were analysed using generalised linear mixed models (GLMMs). Semi-parametric permutational multivariate analysis of variance w..., We selected 13 western Amazonian forest regions ranging from Ecuador to Bolivia. Regions were located throughout a 1,800-km latitudinal gradient. The forest regions encompassed one or two different forest types comprising floodplains, terra firme, and submontane forests. In each region, six to 14 0.1-ha plots were established according to a standard sampling protocol. Plots (50 × 20 m) were located at least 300 m apart, and areas affected by human and natural disturbances were avoided. Overall, we sampled 118 plots in floodplain (n = 27), terra firme (n = 51), and submontane (n = 40) forests. In each plot, we measured and inventoried all individual woody plants (trees, palms, tree ferns, and lianas) with DBH ≥ 2.5 cm rooted within the plot limits. The 118 0.1-ha plots contained 31,20 individual woody plants with DBH ≥ 2.5 cm, and 1,978 species that belonged to 557 genera and 112 families. , Taxonomic standardisation of species names was based on the Plant List using the R package ‘Taxonstand’.