『Abstract
　 Rice is one of the most important staple foods in the world. Lowland paddy fields are well known for functioning as denitrification areas, but few studies have been conducted of paddy fields situated on hill slopes (terraced paddy fields). These terraced paddy fields have a characteristic artificial stepped shape, and this unique shape and periodic ponding from rice production may configure unique hydrological properties that might be different from lowland paddy fields. The shape and hydrological properties may also affect transport of nutrients such as nitrogen.
　This study is particularly focused on the denitrification rate in terraced paddy fields. To understand the hydrological properties of terraced paddy fields, a detailed water budget including the subsurface flow components was calculated. Combining the water budget components and chemical measurements of surface and subsurface water, a nitrogen budget was calculated. The results showed that about 10% of the total nitrogen input, mainly from fertilizers, was lost, suggesting the occurrence of denitrification in the area. The average denitrification rate of the study site was estimated at about 0.53-0.67 g N m^-2 year^-1.Spatial variations in the measured groundwater nitrate concentration suggest that denitrification is important in both the plough layer and the sloping area. The denitrification rate in the sloping area was estimated at 0.67-0.78 g N m^-2 year^-1, which is slightly higher than the estimates of denitrification rate in paddy lots, i.e., 0.56-0.61 g N m^-2 year^-1. The result indicates the importance of sloping areas for denitrification in terraced paddy fields.

Keywords: Terraced paddy; Wetland; Subsurface flow; Slope; Denitrification rate』

1. Introduction
2. Materials and methods
　2.1. Study site
　2.2. Observations
　2.3. Geological setting
　2.4. Nitrogen and nitrate budget calculation
3. Results and discussion
　3.1. Characteristics of subsurface water flow
　3.2. Water budget
　3.3. General characteristics of chemical components and nitrogen behavior
　3.4. Estimate of denitrification rate by nitrogen and nitrate budget
4. Conclusions
Acknowledgments
Appendix A
References