Aiming to address the challenges of poor nitrogen (N) removal and greenhouse gas (GHG) emissions in aquaculture wastewater treatment, biological-electrolysis constructed wetland (BECW) and constructed wetland (CW) were established to comparatively evaluate the N removal efficiency, the greenhouse gas (GHG) emissions, and specific microbial gene abundances under various influent C/N ratios. Increasing influent C/N ratio markedly enhanced the removal rates of COD, NH4+-N, NO3−-N, and TN, while markedly reduced N2O fluxes in CW and BECW. The integration of an electric field with CW (BECW) significantly compensated for the shortcomings of the poor NO3−-N removal efficiency and high N2O emissions when C/N < 6. However, this compensating effect gradually weakened with the continue increase of C/N. BECW simultaneously achieved the high COD (89.26 ± 6.08 %), NO3−N (97.03 ± 1.22 %), and TN (83.17 ± 3.19 %) removal rates, and low GHG emissions (486.00 ± 36.61 mg/m2/h) at C/N = 6. The BECW significantly decreased the abundance of mcrA and pmoA genes while increased the abundance of nirK, nirS, and nosZ genes, and the ratios of pmoA/mcrA and nosZ/(nirS + nirK) across all C/N condition. The emission flux of CH4 in BECW was positively correlated with C/N and the mcrA abundance but significantly negatively correlated with pmoA abundance and pmoA/mcrA. The emission flux of N2O was positively correlated with nirS and nirK abundance but significantly negatively correlated with C/N, NO3−-N removal rates, nosZ abundance, and the index of nosZ/(nirS + nirK). This research offers an innovative approach to get both low GHG emissions and high NO3−-N removal rate in aquaculture wastewater treatment.