Trade-off mechanisms between N2O emissions and nitrogen multifunctionality in a lake littoral mesocosm wetland under seasonal water-level fluctuations: insights from metagenomics and machine learning.

Freshwater littoral wetlands are hydrologically dynamic interfaces that regulate nitrogen (N) metabolism (e.g., removal, retention, and N2O emission); however, the effect of seasonal water-level fluctuations on N2O emissions and nitrogen multifunctionality (NMF) remains poorly quantified. This hydrological variability challenges accurate estimates of greenhouse gas (GHG) emissions and complicates ecosystem management strategies that aim to balance climate mitigation with ecosystem functions and sustainability. Here, using a water-level controlled mesocosm at Poyang Lake Wetland Research Station, China, we combine static-chamber measurements, 15N isotope pairing, metagenomic binning, and an interpretable causal machine learning framework to elucidate the microbial mechanisms and environmental thresholds governing the trade-off between limiting N2O emissions and maintaining NMF. N2O flux (-27.678 to 86.791 μg m-2 h-1) was observed at the source-sink transition with rising water levels, whereas NMF was higher in both the continuously dry (0.379) and wet (0.158) zones than in zones subject to water-level fluctuations. A functional quadrant plot revealed the asynchronous relationship between N2O emissions and NMF maintenance. Metagenomic binning demonstrated that distinct dominant microbial taxa mediated N2O and NMF and their trade- off via cooperative and competitive interactions. Moreover, key thresholds, including soil organic matter contents and abundances of hao, hzsABC, nosZII, nirKS, and nasAB genes, drive the system toward a low-emissions and high-function state. This study clarifies the trade-off mechanisms between N2O emissions and NMF maintenance, and provides an ecological basis for reconciling climate mitigation with ecosystem functions in aquatic ecosystems.