Seasonal nitrous oxide emission from different land use in tropical riparian ecosystem : a case study in Nan province, Northern Thailand

Abstract

One of an important ecological services provided by tropical riparian ecosystems is the mitigating contamination of ammonium, nitrate, and phosphate leaching from agricultural area to water resources. However, a negative impact of this pollutant remediation may be that the ecozone also functions as a source of nitrous oxide (N2O) emission. The objectives of this study were to measure the N2O emission in such an ecosystem with specific emphasis on temporal aspects; comparing between wet and dry seasons and different land use; comparing a leguminous reforestation, Samanea saman (Jacq.) Merr, with applied nitrogen fertilizer in conventional agriculture with maize, Zea mays L, and to identify the major drivers controlling these emission. The results revealed that the annual average emission of N2O from the reforestation (3.3 kg N2O-N ha-1 y-1) was significantly higher than the agricultural areas with maize (2.2 kg N2O-N ha-1 y-1) (P< 0.05). The rate of N2O flux in the wet season was higher than in the dry season (P< 0.05). The N2O emission variability was correlated with the controlling factors; water filled pore space (WFPS), denitrification, and microbial biomass carbon. When inorganic nitrogen and soil organic carbon are sufficient, WFPS plays an important role in controlling N2O emission contributed by denitrification. N2O flux observed by distal proximity to river in a mixture transect was significantly different both wet and dry seasons (P< 0.05) in that N2O flux increased where sampling locations were closed to river. This pattern was correlated to the slowly decreasing amounts of inorganic nitrogen and dissolved organic carbon from upper agricultural field boundary to lower river side. Conversely, WFPS and denitrification increased in the opposite patterns of those relationships. N2O flux from maize area simulated by the DeNitrification-DeComposition model (DNDC) was underestimated when validated with those observed from filed experiment. Sensitivity analysis indicated that N2O emission variability by DNDC model was dependent on soil organic carbon, WFPS, and nitrogen input to maize plot, respectively. Comparatively, annual N2O emission from the reforestation in the tropical riparian zone was similar to those reported for temperate riparian zones and other ecosystems. Although the annual N2O flux from the agricultural area with maize is comparable to other riparian ecosystems, it is higher than those of other N2O flux from terrestrial zones. The results suggest that tropical riparian ecosystem surrounding agricultural land does not represent a major hotspot of N2O flux and does not diminish the positive benefits which they provide in relation to other aspects of ecosystem service provision.