Urban streams are severely impacted by anthropogenic pressures, resulting in reduced water quality and habitat diversity, which leads to a decline of benthic invertebrate diversity. Stream restoration aims to counteract this biodiversity decline by improving instream conditions and increasing habitat diversity, thereby creating new habitats for benthic invertebrates. However, benthic community recovery is often slow or absent. Reasons for this may be found in the mechanisms that control the recolonisation process: dispersal, tolerances (environmental filter) and biotic interactions (e.g., competition). According to the ‘Asymmetric Response Concept’ (ARC, Vos et al., 2023) their roles change predictably in space and time, but this has never been tested. To understand how dispersal, tolerances and competition interact during recolonisation, the three mechanisms need to be investigated in concert and their relative importance considered. Existing studies often lack detailed long-term data, leaving the patterns of recolonisation unresolved. This thesis addresses this research gap, using time-series data from the Boye catchment (Western Germany), which is characterised by urban land use and was used as open sewer for almost a century. The removal of wastewater, combined with hydromorphological restoration initiated a recolonisation process. This process and the underlying mechanisms were addressed in three chapters.

In the first chapter, ten years of monitoring data were analysed to identify general recolonisation patterns of benthic invertebrates and corresponding environmental drivers (Gillmann et al., 2023). Benthic communities became more similar and stabilised after eight years, with time since restoration being the main driver of community development. Habitat development, coupled with the increase in woody riparian cover, led to a shift in species assemblages. In the second chapter, the same time-series data was used to analyse the temporal changes of the communities’ traits, coupled to dispersal, tolerance and competition (Gillmann et al., 2024a). The communities’ dispersal capacity was highest directly after restoration and gradually decreased over time. Simultaneously, tolerance toward organic decomposition decreased, while tolerance toward chloride remained stable. On the other hand, interspecific competition (i.e., trait overlap) increased with time since restoration. In the last chapter, the role of the three filters was investigated based on 48 samples, collected in the Boye catchment in 2022. For the analysis, species- and site- specific data was used, including distance to source populations (dispersal filter), habitat suitability (environmental filter) and trait overlap (competition filter) (Gillmann et al., 2024b). Similar to chapter two, the dispersal filter was identified to be most important during initial recovery, while it continued to influence species distributions at more mature sites. The environmental filter significantly affected species distributions at sites unimpacted by historical sewage input and hydromorphological restoration. In contrast, a moderate effect was observed at ‘mature restored’ (> 10 years) sites, while no effect was observed at ‘recently restored’ (< 4 years) sites. Lastly, the competition filter was not a determining factor for species distributions, regardless of the restoration age.

The findings from this thesis are in line with the major assumption of the ARC that initial recovery depends on dispersal (Vos et al., 2023). Tolerances remain important if stressors are removed gradually. On the other hand, natural succession results in the increasing importance of environmental conditions over time. The competition filter showed contradicting results, indicating that trait overlap does not result in competitive exclusion.

In conclusion, this thesis provides valuable insights into the patterns and mechanisms acting during recolonisation of benthic invertebrates in urban streams. Despite the open sewer history of the Boye catchment, the communities recovered, following comparable patterns over time. This emphasizes that stream restoration efforts should continue to be directed at urban streams. These should consider the recovery trajectory of the ARC to remove possible barriers to recolonisation.