Mechanisms of crop management impact on methane emissions from rice fields in Los Banos, Philippines

This article comprises 4 yr of field experiments on methane (CH(4)) emissions from rice fields conducted at Los Banos, Philippines. The experimental layout allowed automated measurements of CH(4) emissions as affected by water regime, soil amendments (mineral and organic), and cultivars. In addition to emission records over 24 h, ebullition and dissolved CH(4) in soil solution were recorded in weekly intervals. Emission rates varied in a very wide range from 5 to 634 kg CH(4) ha(-1), depending on season and crop management. In the 1994 and 1996 experiments, field drying at midtillering reduced CH(4) emissions by 15-80% as compared with continuous flooding, without a significant effect on grain yield. The net impact of midtillering drainage was diminished when (i) rainfall was strong during the drainage period and (ii) emissions were suppressed by very low levels of organic substrate in the soil. Five cultivars were tested in the 1995 dry and wet season. The cultivar IR72 gave higher CH(4) emissions than the other cultivars including the new plant type (IR65597) with an enhanced yield potential. Incorporation of rice straw into the soil resulted in an early peak of CH(4) emission rates. About 66% of the total seasonal emission from rice straw-treated plots was emitted during the vegetative stage. Methane fluxes generated from the application of straw were 34 times higher than those generated with the use of urea. Application of green manure (Sesbania rostrata) gave only threefold increase in emission as compared with urea-treated plots. Application of ammonium sulfate significantly reduced seasonal emission as compared with urea application. Correlation between emissions and combined dissolved CH(4) concentrations (from 0 to 20 cm) gave a significant R(2) of 0.95 (urea + rice straw), and 0.93 (urea + Sesbania), whereas correlation with dissolved CH(4) in the inorganically fertilized soils was inconsistent. A highly significant correlation (R(2) =0.93) existed between emission and ebullition from plots treated with rice straw. These findings may stimulate further development of diagnostic tools for easy and reliable determination of CH(4) emission potentials under different crop management practices.