Effects of 40-60 % reduction in fertilizer use by single application in a nursery box and non-puddling cultivation on outflow loads from paddy fields, mass balance and hulled rice yields were examined for six years. Four experimental plots were prepared for comparison: conventional puddling and conventional fertilizer use, conventional puddling and reduced fertilizer use, non-puddling and conventional fertilizer use, and non-puddling and reduced fertilizer use.The levels of surface effluent loads of the plots with single application of fertilizer in a nursery box were lower by 43 % for total nitrogen (T-N) and 39 % for total phosphorous (T-P) compared to plots in which conventional fertilizer application was used. Furthermore, subsurface effluent loads were 67 % lower for dissolved nitrogen (D-N) and were dependent upon the fertilizer reduction rates for dissolved phosphorous (D-P). Suspended solids (SS), chemical oxygen demand (COD), T-N, and T-P surface runoff loads from the non-puddling divisions were lower by 41 %, 46 %,59 %, and 65 %, respectively, compared to those from puddling treatments. In the conventional fertilizer treatments, both N and P collected from input factors such as irrigation, rain and fertilizer use were higher than those collected from output factors such as surface effluent load, subsurface effluent load, and hulled rice, because of the increased fertilizer use, which exceeded 11 % for N and 44 % for P. In the single-application fertilizer treatments, the hulled rice yields were 6 % lower,but the rates of yield to the amount of fertilizer were higher.

View full abstract

 To investigate regional long-term trends in annual precipitation from a limited number of observation stations, it is useful to understand the relation between the spatial representation of data and the number of years of long-term data examined. As a case study, we used 79 years of precipitation data from five stations near Nagoya, Japan, to analyze the relation between the spatial representation and number of years of data. When the years of data were fewer than 45, the trend differed for all five stations. However, when more than 70 years of data were used, the trend was the same for all five stations. When the years of data were more than 45 but fewer than 70, four of the five stations exhibited the same trend. Results show that at least 70 years of data are necessary to determine the long-term trend of precipitation in our study area, which is a large watershed affected by both Pacific climate and Sea of Japan climate.

View full abstract

This article published in J. Japan Soc. Hydrol. and Water Resour., 2012, 25(3), 152-162 was withdrawn by the Editorial-publishing committee on June 4, 2013, due to inaccuracy in its writing.

View full abstract