Biodiversity is structured by multiple mechanisms that are dependent, at least in part, on ecological similarities and differences among species. Integrating traits and phylogenies in diversity metrics may provide deeper insight into community assembly processes across spatial scales. However, different traits are influenced by processes at different spatial scales, and it is not clear how trait‐spatial scale mismatches skew our ability to detect assembly patterns. An additional complexity is how phylogenetic distances, which might capture unmeasured traits, reflect spatially dependent processes. Here we analyze a freshwater zooplankton dataset from 91 ponds and show that different traits are associated with processes at different spatial scales. We first assessed the response of individual traits to processes at both α‐ and β‐scales, and then quantified the power of different combinations of traits and phylogenetic distances to reveal environmental and spatial drivers of α‐ and β‐diversity. We found that explanatory power was maximised when we accounted for environmental and spatial drivers with single, but different traits for α‐ and β‐diversity. Using the most appropriate trait for each spatial scale outperformed phylogenetic information, but phylogenetic information outperformed the same traits when these were used at the wrong spatial scale, and all outperformed taxonomic analyses that ignore trait and phylogenetic information. We demonstrate that accounting for species’ similarities and differences provides important information about dominant assembly mechanisms at different spatial scales, and that phylogeny is especially useful when measured traits are uninformative at a given spatial scale or when there is lack of trait data. Our study also indicates, however, that trait‐scale mismatches among phylogenetically conserved traits may affect the performance of phylogenetic indices compared to indices that account only for the best single trait at each spatial scale.