Biological diversity on islands is shaped by an interplay of geo-environmental dynamics that are relatively easy to observe, as islands have distinct boundaries and are often small in size. These unique characteristics of islands have inspired naturalist to their use as research laboratories. Islands have therefore influenced many theories in ecology, evolution, and biogeography. Despite great scientific advances in island research, however, we still lack comprehensive understanding of numerous aspects in island ecology. Open research questions are related to the effects of spatial scales on biodiversity, community assembly processes, and diversity of species forms and functions on islands.The central aim of my thesis, which consists of five research chapters that consecutively build on each other, was to investigate species diversity patterns and assembly processes on small islands. In chapter 2, I present a new method to measure leaf area, which was crucial for the plant trait analyses. In chapters 3-5, I sampled and analysed a dataset of tree communities on 60 small islands in the Raja Ampat Archipelago to gain detailed ecological knowledge of local patterns and processes of species diversity. In chapter 6, I compiled a global dataset that contains information on plant species occurrences and their respective growth forms for more than 5100 species on 700 small islands to test for global patterns in growth form species richness.Based on the Raja Ampat dataset, I showed that the species-area relationship strongly depends on the applied sampling scale. In addition, environmental factors that determined species richness differed between scales. Species occurring on these islands have specific environmental requirements, are linked to island area and form island-specific species pools. Species communities on smaller islands had a lower community completeness than those on larger ones. This observation was likely rooted in local limiting processes, acting more severely on smaller islands and thus preventing many species of the pool to establish. These local limiting processes can be attributed to a combination of nonrandom dispersal and environmental filters operating simultaneously, but with different intensity on island communities. Species communities on smaller islands were shaped by dispersal constraints and limited niche space. In contrast, communities on larger islands were governed by niche partitioning related to nutrient, light, and resource acquisition strategies. At a global scale, species richness patterns dissected into growth forms varied