A short-term investigation of trace gas emissions following tillage and no-tillage of agroforestry residues in western Kenya

Improved-fallow agroforestry systems are increasingly being adopted in the humid tropics for soil fertility management. However, there is little information on trace gas emissions after residue application in these systems, or on the effect of tillage practice on emissions from tropical agricultural systems. Here, we report a short-term experiment in which the effects of tillage practice (no-tillage versus tillage to 15 cm depth) and residue quality on emissions of N2O, CO2 and CH4 were determined in an improved-fallow agroforestry system in western Kenya. Emissions were increased following tillage of Tephrosia candida (2.1 g N2O-N ha−1 kg N applied−1; 759 kg CO2-C ha−1 t C applied−1; 30 g CH4-C ha−1 t C applied−1) and Crotalaria paulina residues (2.8 g N2O-N ha−1 kg N applied−1; 967 kg CO2-C ha−1 t C applied−1; 146 g CH4-C ha−1 t C applied−1) and were higher than from tillage of natural-fallow residues (1.0 g N2O-N ha−1 kg N applied−1; 432 kg CO2-C ha−1 t C applied−1; 14.7 g CH4-C ha−1 t C applied−1) or from continuous maize cropping systems. Emissions from these fallow treatments were positively correlated with residue N content (r = 0.62–0.97; P < 0.05) and negatively correlated with residue lignin content (r = −0.56, N2O; r = −0.92, CH4; P < 0.05). No-tillage of surface applied Tephrosia residues lowered the total N2O and CO2 emitted over 99 days by 0.33 g N2O-N ha−1 kg N applied−1 and 124 kg CO2-C ha−1 t C applied−1, respectively; estimated to provide a reduction in global warming potential of 41 g CO2 equivalents. However, emissions were increased from this treatment over the first 2 weeks. The responses to tillage practice and residue quality reported here need to be verified in longer term experiments before they can be used to suggest mitigation strategies appropriate for all three greenhouse gases.