Archaeal communities in arable soils are dominated by Nitrososphaeria, a class within Thaumarchaeota comprising all known ammonia-oxidizing archaea (AOA). AOA are key players in the nitrogen cycle and defining their niche specialization can help predicting effects of environmental change on these communities. However, hierarchical effects of environmental filters on AOA and the delineation of niche preferences of nitrososphaerial lineages remain poorly understood. We used phylogenetic information at fine scale and machine learning approaches to identify climatic, edaphic and geomorphological drivers of Nitrososphaeria and other archaea along a 3000 km European gradient. Only limited insights into the ecology of the low-abundant archaeal classes could be inferred, but our analyses underlined the multifactorial nature of niche differentiation within Nitrososphaeria. Mean annual temperature, C:N ratio and pH were the best predictors of their diversity, evenness and distribution. Thresholds in the predictions could be defined for C:N ratio and cation exchange capacity. Furthermore, multiple, independent and recent specializations to soil pH were detected in the Nitrososphaeria phylogeny. The coexistence of widespread ecophysiological differences between closely related soil Nitrososphaeria highlights that their ecology is best studied at fine phylogenetic scale.