poster - International Conference of Agricultural Engineering

Land use in each sub-watershed was characterized using GIS as rice paddy fields, forest, and
upland fields with a ratio of each land use area to the total area of each sub-watershed.
Concentration of nitrate dissolved in sample waters was quantitatively determined by an ion
chromatography using Dionix IC 20. Two litters of the sample waters were passed through a
Millipore filter with the pore size of 0.2 µm, and concentrated down to about 40 ml on an
electric hot plate, reducing the volume in a stainless steel vat. Special care was paid for
prompt contamination during the concentration procedure. Nitrate in the concentrates
containing nitrogen less than 300 µg was reduced by addition of Devarda’s alloy under the
gas-tight container (Kjeldahl steam distillation apparatus) into ammonia. The evolved
ammonia was quantitatively fixed as an ammonium sulfate in an excess of sulfuric acid
medium. Ammonium in the solution was eventually converted into ammonium salt of
tetraphenyl-borate that was hardly soluble in acidic water. The precipitate was recovered
and air-dried. The 15 N/ 14 N ratios of the dry precipitate were determined by a continuous-flow
mass spectrometer (Finnigan DELTA plus). Nitrogen isotopic ratios were expressed by
common δ 15 N notation, per mil variation relative to atmospheric dinitrogen (δ 15 N = 0 ‰). The
overall precision during the protocol established in the present study is better than ±0.2 ‰.
quantitative recovery of reagent-grade nitrate in range from 300 to 1,000 µg-nitrogen during
the reduction and precipitation, together with the nitrogen isotopic measurement were
repeatedly tested, confirming analytical accuracy.
2.1. Relationship between δ 15 N and NO 3 concentration
As NO 3 in the main stream was mixed with NO 3 in tributaries, the following relationship for the
main stream might be described with a mixing model (Mariottiet al., 1988):
δ m
Q m
= ∑ δ i
Q i
[1]
i
Q = A×
H × C
[2]
where δ was nitrogen isotopic ratio (‰), Q was the mass of nitrate (kg), A was the area of a
watershed (km 2 ), H was precipitation (mm), and C was NO 3 concentration of river water
(mg/L). Subscripts m and I indicated the main stream and a tributary, respectively. If we
assumed that precipitation in all sub-watersheds was identical, Eqs. 1 and 2 were reduced to:
δ C
2.2. Estimation of NO 3 origin
m
m
=
∑
i
δ A C
i
i
i
A
m
. [3]
The amount of N supplied to river water might be estimated based on differences in land use
and different δ 15 N values for the various origins of N, i.e. chemical fertilizer, animal manure
compost, and rain water. There were following relationships of NO 3 concentration and δ 15 N
for those origins (Morita, 2002):
A=B+C+D [4]
aA=bB+cC+dD [5]
where A, B, C, and D indicated the NO 3 concentrations (mg/L) of river water, chemical
fertilizer origin, animal manure origin, and rainwater origin, respectively, and a, b, c, and d
indicated δ 15 N values (‰) of NO 3 for river water, chemical fertilizer origin, animal manure
origin, and rainwater origin, respectively.
The NO 3 concentration of the chemical fertilizer origin, B (mg/L), might be estimated by
substituting measured values of A and a for Eqs. 4 and 5, respectively, as: