.png){kind=icon}
.png){kind=icon}
Return
Enhancing urban river self-purification through riverbed substrates configuration: A nature-based solution for nutrient removal and restoration planning
.png){kind=icon}
PRE
AI
Y
L
S
Q
Y
DOI：10.1016/j.jes.2026.01.038.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Appropriate riverbed substrates, as nature-based engineering components, are critical for enhancing nutrient mitigation and ecosystem sustainability in urban rivers. However, their role in regulating hydrologically mediated nutrient fluxes and biofilm functions remains unclear, limiting substrate-optimized design for urban river restoration. This study integrated machine learning modeling, scenario simulations, and metagenomic analysis to quantify substrate-driven interfacial nutrient removal efficiencies and uncover microbial regulation mechanisms. A back propagation neural network could accurately predict interfacial ammonium and total organic carbon removal efficiencies (RMSE: 0.59-6.92 mg/(L & centerdot;h & centerdot;m(2)), R-2: 0.66-0.97), with retention time, temperature, dissolved oxygen, and nutrient load identified as key predictors. Building upon the model-predicted scenario results, analysis of similarity tests confirmed that substrate type significantly influenced interfacial nutrient removal efficiencies (R > 0.05, P < 0.001). Scoring metrics demonstrated fine sand (1295) and gravel (1281) gained higher total scores than other substrates (1110-1182), indicating higher interfacial nutrient removal capacities. Metagenomic analyses revealed that these differences were driven by divergence in microbial functional potential. Substrate type selectively enriched functional genes related to nitrogen and carbon cycling (R > 0.18, P < 0.05), with gravel microcosms showing significantly higher gene abundance (8.00 & times; 10(-4)-2.08 & times; 10(-3)), despite similar community compositions governed by stochastic assembly (R & sup2; > 0.84). Topological analysis revealed that redundancy of functional gene network significantly influenced ammonium removal efficiency (P < 0.05), with fine sand and gravel enhancing ammonium removal, while lower clustering coefficients in artificial filler and gravel microcosms significantly promoted total organic carbon removal. This study suggested that fine sand and gravel should be more effective riverbed substrates for enhancing interfacial nutrient removal in urban river restoration.
Keywords:
Riverbed substrate
River restoration
Interfacial nutrient removal
Microbial metabolism
Community assembly
Journal
IF:
6.3
Papers:
7.5K
Citations:
2.4W
Organization
Cited Papers