ABSTRACT Water management impacts both methane (CH4) and nitrous oxide (N2O) emissions from rice paddy fields. Although controlled irrigation is one of the most important tools for reducing CH4emission in rice production systems it can also increase N2O emissions and reduce crop yields. Over three years, CH4 and N2O emissions were measured in a rice field in Uruguay under two different irrigation management systems, using static closed chambers: conventional water management (continuous flooding after 30 days of emergence, CF30); and an alternative system (controlled deficit irrigation allowing for wetting and drying, AWDI). AWDI showed mean cumulative CH4 emission values of 98.4 kg CH4 ha−1, 55 % lower compared to CF30, while no differences in nitrous oxide emissions were observed between treatments ( p > 0.05). No yield differences between irrigation systems were observed in two of the rice seasons ( p > 0.05) while AWDI promoted yield reduction in one of the seasons ( p< 0.05). When rice yield and greenhouse gases (GHG) emissions were considered together, the AWDI irrigation system allowed for lower yield-scaled total global warming potential (GWP). Higher irrigation water productivity was achieved under AWDI in two of the three rice seasons. These findings suggest that AWDI could be an option for reducing GHG emissions and increasing irrigation water productivity. However, AWDI may compromise grain yield in certain years, reflecting the importance of the need for fine tuning of this irrigation strategy and an assessment of the overall tradeoff between relationships in order to promote its adoption by farmers.

Los arrozales son fuente de dos importantes gases de efecto invernadero (GEI), metano y óxido nitroso. Como un paso inicial hacia la obtención de información local, se midieron las emisiones de CH4 y N2O del suelo y de las plantas de arroz mediante la técnica de la cámara estática en experimentos en invernáculo y a campo en el este de Uruguay. En el experimento en invernáculo, se estudió el efecto del momento de inundación (21 y 45 días después de la emergencia) y de la fertilización nitrogenada (0 y 50 kg N ha-1) sobre las emisiones. La inundación temprana y la fertilización nitrogenada tendieron a aumentar las emisiones de N2O. En el experimento a campo, se estudió el efecto de la cobertura invernal y de la fertilización nitrogenada (0 y 82 kg N ha-1). Se detectaron mayores flujos de CH4 durante la etapa reproductiva de la planta en el tratamiento fertilizado con cobertura invernal previa de raigrás. El flujo de N2O fue máximo después de los baños. Los resultados indican que el uso del cultivo de cobertura podría incrementar las emisiones de GEI durante el ciclo del arroz. A pesar de las distintas prácticas de manejo del cultivo empleadas en Uruguay, los flujos de CH4 y N2O se encuentran dentro de los valores informados previamente para arrozales de otras partes del mundo.

Irrigated rice fields are major sources of two important greenhouse gases (GHG), methane and nitrous oxide. As an initial step towards obtaining local information, emissions of CH4and N2O from rice paddy soil were measured by the static chamber technique in greenhouse and field experiments conducted in eastern Uruguay. In the greenhouse experiment, the effect of two flooding moments (21 and 45 days after emergency) and nitrogen fertilization (0 and 50 kg N ha-1) on gas emissions was studied. Early flooding and nitrogen fertilization tended to increase N2O emissions. In the field experiment, effect of winter soil cover crop and nitrogen fertilization (0 and 82 kg N ha-1) were tested. Higher CH4fluxes were observed mainly during the reproductive stage of the plant in the N-fertilized treatment with ryegrass winter crop. N2O flux peaked at flushing. Results indicate that the use of cover crops might increase GHG emissions during the rice cycle. Despite differences in agronomic management practices employed in Uruguay, CH4and N2O fluxes are within magnitudes previously reported for rice fields worldwide.

The bacterial communities in floodwater, from a rice-planted and an unplanted field were characterized at the beginning (flooding stage) and at the end (harvest stage) of the rice cropping cycle. Most probable number estimations and plate counts of aerobic and anaerobic heterotrophic bacteria and of several metabolic bacterial groups (methanogens, sulfate-reducers, anaerobic sulfur and nonsulfur phototrophs, denitrifiers and ammonifiers) were similar in rice and unplanted floodwater at both sampling times. The analysis of denitrifiers and methanogens by fluorescent in situ hybridization revealed a shift in the phylogenetic affiliation only of the former group in the rice-planted floodwater. Terminal restriction fragment length polymorphism of 16S rRNA gene amplicons indicated that the bacterial communities of the rice-planted and unplanted soils were consistently diverse and strongly influenced by the season.