Degradation and recovery trajectories

Degradation and recovery trajectories

Our results highlight the variability of degradation and recovery trajectories, thus specifying the original Main Hypotheses, specifically MH1. We found evidence for that variability for various ecosystem components:

  • Gillmann et al. (2023) observed an increasing similarity of benthic invertebrate communities with time since restoration in the Boye catchment, indicating a random starting community that is increasingly streamlined with the development of environmental conditions.
  • With benthic invertebrates collected over 12 years in the recovering Boye catchment, stable isotope analyses were performed (see Schlenker et al., 2024), which is the first analysis of the long-term recovery of food webs. Consistently over all sites, trophic similarity increases with time since restoration suggesting that the stability and redundancy of the food webs increases with recovery.
  • A leaf-litter decomposition experiment at 20 stream sites within the Boye catchment tested the hypothesis that litter decomposition would be least efficient at recently restored sites (David et al., in prep). However, decomposition dynamics was independent of the time elapsed since restoration and mainly explained by land use, pH, and chloride concentrations, suggesting extremely rapid recovery of ecosystem functioning or persisting functional impairment across the stream network.
  • Traits better reflect recovery processes than compositional metrics. In the only slightly disturbed Kinzig catchment, community development over time was less pronounced. Spatially, consistent patterns over a decade of sampling revealed that less disturbed communities were characterised by higher taxonomic and trait diversity and occurrence of pollution-sensitive taxa. Taxonomic/trait metrics consistently reflected the magnitude of the environmental improvement, while community composition did not, supporting the hypothesis that functions recover faster than communities (see Nguyen et al., 2023).