Chemical composition, or quality, of agroforestry residues influences N2O emissions after their addition to soil

Emissions of N2O were measured following addition of 15N-labelled (2.6–4.7 atom% excess 15N) agroforestry residues (Sesbania sesban, mixed Sesbania/Macroptilium atropurpureum, Crotalaria grahamiana and Calliandra calothyrsus) to a Kenyan oxisol at a rate of 100 mg N kg soil−1 under controlled environment conditions. Emissions were increased following addition of residues, with 22.6 mg N m−2 (124.4 mg N m−2 kg biomass−1; 1.1 mg 15N m−2; 1.03% of 15N applied) emitted as N2O over 29 d after addition of both Sesbania and Macroptilium residues in the mixed treatment. Fluxes of N2O were positively correlated with CO2 fluxes, and N2O emissions and available soil N were negatively correlated with residue lignin content (r=−0.49;P<0.05), polyphenol content (r=−0.94;P<0.05), protein binding capacity (r=−0.92;P<0.05) and with (lignin+polyphenol)-to-N ratio (r=−0.55;P<0.05). Lower emission (13.6 mg N m−2 over 29 d; 94.5 mg N m−2 kg biomass−1; 0.6 mg 15N m−2; 0.29% of 15N applied) after addition of Calliandra residue was attributed to the high polyphenol content (7.4%) and high polyphenol protein binding capacity (383 μg BSA mg plant−1) of this residue binding to plant protein and reducing its availability for microbial attack, despite the residue having a N content of 2.9%. Our results indicate that residue chemical composition, or quality, needs to be considered when proposing mitigation strategies to reduce N2O emissions from systems relying on incorporation of plant biomass, e.g. improved-fallow agroforestry systems, and that this consideration should extend beyond the C-to-N ratio of the residue to include polyphenol content and their protein binding capacity.