PAWEES 2013 Full Paper

Contents
Session 1
A-02 Impact of Average Elevation of Basins on Earlier Snowmelt Caused by Climate Change
ITO Nobuo, SUTO Yuji, NAKAMURA Kazumasa ································································································ 1
A-03 Assessment of Climate Change Impact on Crop Yields in Northern Taiwan by Using
Principal Component Analysis
RAY-SHYAN WU, MING-HSU LI, JI-TANG FANG, CHI-MEI WANG ···························································· 9
A-06 Climate Change Action Plan for Water Resources in Taiwan

Wei-Fu Yang, Chi-Ming Chen, Pei-Jung Wu ······································································································ 21
Session 2
B-04 Water-saving Effect of Simplify Surge Flow Method ADF Method in Uzbekistan
Junya Onishi, Paluashova Ghavharay, Hiroshi Ikeura ····················································································· 31
B-07 Expansion of Leased Paddy Land and Crisis of Sustainability of Water User Associations in Japan
Hajime Tanji, Katsuhiro Sakurai, Ataru Nakamura, Hirohide Kiri, Tetsuo Nakaya ···························· 43
Session 3
C-01 Water Management at Large-Sized, Sub-Irrigation-Installed Paddy Fields
Nakamura Kazumasa, Kohiyama Masayuki, Unoki Keiji ··············································································· 53
C-03 Development of a Simple Method of Discrimination between the Dojo and Kara-dojo Loaches
for the Conservation of Japan’s Rural Ecosystem
Noriyuki Koizumi, Kazuya Nishid, Atsushi Mori, Keiji Watabe, Takeshi Takemura ·························· 67
C-07 Determining Optimal Soil Moisture for Irrigated Rice in Indonesia with System of
Rice Intensification
Chusnul Arif, Masaru Mizoguchi, Budi Indra Setiawan, Tsugihiro Watanabe ···································· 75
[1]
Session 4
D-01 Decrease of Egg-masses for the Japanese Brown Frog (Rana japonica) after Land
Consolidation Project in Paddy Field Area, Japan
Keiji Watabe, Atsushi Mori, Noriyuki Koizumi, Takeshi Takemura, Kazuya Nishida ····················· 87
D-02 Applicability Study of Ecological Impact Assessment Using AQUATOX Model in Paldang
Reservoir, South Korea
Chun Gyeong Yoona, Han-Pil Rheeb, YeongKwon Sonc ·········································································· 93
D-05 Feasibility Analysis of Nitrogen Balance in Paddy Fields toward New Irrigation Service for
Rice Quality
Tasuku Kato, Toshiaki Iida ····································································································································· 111
D-07 Relational Analysis between Yield and Planting Condition of Rainy Season Rice in Low
Productive Fields:a Case Study in Lao PDR
Hiroshi Ikeura, Phetyasone Xaypanya, Sengthong Phongchanmixay, Somphone Inkhamseng,
Somnuck Soubat, Salermphon Phonangeone, Soulintha Chanthabuly ············································· 115
D-08 Investigation of Organic Fertilizer to Reduce Insecticide - Assessment of Paddy Ecosystem

using Emergence Husks of Red-Dragonflies -
Aoda Tadao, Katano Kai, Toyama Kazunari, Jinguji Hiroshi ···································································· 123
Session 5
E-01 Screening Rice (Oryza sativa L.) Varieties Suitable for System of Rice Intensification (SRI)
K. Noborio, J. Lanceras-Siangliw, K. Katano, M. Mizoguchi, T. Toojinda ··········································· 129
E-02 Effect of SRI Methods on Water Use, NPS Pollution Discharge, and GHG Emission in Korean Trials
Joongdae Choi, Gunyeob Kim, Woonji Park, Suin Lee, Deogbae Lee, Dongkoun Yun ·············· 133
E-04 The Impact of Agriculture Policy to Rural Water Management in Northern Taiwan
Ray-Shyan Wu, Chia-Chi Ma ································································································································· 145
E-05 Irrigation Practice and Irrigation Management Improvement in Baingda Irrigation Project
Maung Maung Naing, Thiha Aung, Zaw Min Htut, Yutaka Matsuno, Haruhiko Horino ············ 155
E-06 Nitrogen and Weed Management in No-tilled Transplanted Rice on No-tilled Transplanted
Rice- Surface Seeded Wheat Cropping System under Conservation Agriculture
Pijush K Mukherjee, Biswapati Sinha ·················································································································· 163
[2]
Session 6
F-02 Modeling the Future Water Footprint of Paddy Rice in the Republic of Korea
Temba Nkomozepi, Sang-Ok Chung ··············································································································· 175
F-03 Effect of Return Flow on Water Temperature in Irrigation-Drainage Canal Under Spill-over
Paddy Irrigation
Masaomi KIMURA, Kouki KASAI, Toshiaki IIDA, MARIE Mitsuyasu, Naritaka KUBO ·················· 187
F-05 The Suitability Evaluation of Dredged Soil from Reservoir as Embankment Materials
Jaesung Park, Younghwan Son, Sookack Noh, Taeho Bong ·································································· 199
F-08 Analysis of Irrigation Service Needs by Rice Farming Families in Japan

Toshiaki IIDA, Masaomi KIMURA, Koshi YOSHIDA, Naritaka KUBO, Takahiro YOKOI ················ 211
F-10 Runoff Characteristics of Non-point Source Pollution from Reclaimed Paddy

Yujin Lee, Chun Gyeong Yoon, Joon-Sik Kim, Moonsoo Cho, Seungil Lee ····································· 219
F-15 Evaluation of Effects on Baseflow of Using Measured Field Slope Length and Slope using SWAT
Ji Min Lee, Younghun Jung, Gwan Jae Lee, Seong Joon Kim, Joong Dae Choi, Kyoung Jae Lim
······························································································································································································ 229
F-16 Assessment of Paddy Field Runoff on Water Quality of Yeongsan River Basin by Load
Duration Curve
Dongho Choi, Jaewoon Jung, Kwangsik Yoon, Woojung Choi, Hana Park ····································· 235
Poster Session
P-03 An Analysis of Runoff Characteristics of Hosan Stream Using Rainfall-Runoff Model
Seung J. Maeng, Ji H. Shim, Gil S. Hwang, Dong O. Kim, Ji H. Jeong ············································· 243
P-04 Development of Irrigation Management Method for Reducing Inflow of Radioactive

Substances in Japan
Moono Shin, Tomijiro Kubota, Koji Hamada, Tadayoshi Hitomi ························································ 249
P-06 National Risk Assessment of Irrigation on the Farmland near Wastewater Treatment
Plants in Korea
Jae-Ho Choi, Chun Gyeong Yoon, Han-Pil Rhee, Moonsoo Cho, Je ha Ryu ································ 255
[3]
Session 1
PAWEES 2013 (12TH) INTERNATIONAL
CONFERENCE ON AGRICULTURAL
WATER AND RURAL ENVIRONMENT FOR
THE FUTURE
RAMADA PLAZA HOTEL, Cheongju, KOREA
Wednesday, Oct. 30 – Friday, Nov. 1, 2013
A-02
Impact of Average Elevation of Basins on Earlier Snowmelt

Caused by Climate Change
ITO Nobuo, SUTO Yuji, NAKAMURA Kazumasa

Civil Engineering Research Institute for Cold Region ,PWRI
ABSTRACT
In snowy cold regions such as Hokkaido, snowmelt accounts for a significant amount of the irrigation
water for rice paddy fields. Warming in such regions will shift the snow-melting season earlier, which
may alter the water balance in irrigation. The arrival of the snow-melting season and the snowmelt
runoff are thought to vary by elevation. In a single catchment basin, the reservoirs and inlet works at
different elevations may experience varying degrees of change in snow melting and snowmelt runoff.
In light of this, we have investigated how the basin elevation affects the changes in snowmelt runoff,
by using predicted values of temperature and precipitation provided by nine climate models. As a
result, it was predicted that the lower is the average elevation of a basin, the earlier the snowmelt
season will arrive. It was also estimated that the total amount of water that flows into rivers that are
used as water sources will decrease during the snowmelt and irrigation periods. In view of these
findings, irrigation facilities located in a close proximity will need to cooperate to respond to the
earlier snowmelt runoff seasons.
Keywords: Climate change, snow melt runoff, Hokkaido, climate model, water management
1. Introduction
Hokkaido is the northernmost of Japan’s four main islands. The annual mean temperature in
Sapporo, the central city of Hokkaido, is 8.5 °C and annual precipitation there is about 1,100 mm. The
monthly mean temperature from February to December is below 0 °C. Precipitation in winter remains
on the ground as snow. The snow cover on flatlands disappears in March at the earliest and in April at
the latest. In mountainous areas, the thawing season is later.
Arable land in Hokkaido measured 1.15 million ha as of 2012, accounting for about 25% of
Japan’s arable land. 224 thousand ha of the arable land in Hokkaido is paddy field, accounting for
about 9% of Japan’s paddy field area. As these data indicate, Hokkaido is a major food-producing area
of Japan.
Figure 1 shows the role of snowmelt in paddy field irrigation in Hokkaido. Generally, river
discharges here increase from March when snow starts to melt in the mountains. The snowmelt runoff
PAWEES 2013 (12TH) INTERNATIONAL CONFERENCE ON AGRICULTURAL WATER AND RURAL ENVIRONMENT FOR THE FUTURE 1
peaks in the period from April to May. Since rainfall is comparatively lower in June and July, river
discharges decrease during that time. From August, however, rainfall starts to increase, which leads to
increases in river discharges. As shown in figure 1, the demand for paddy field irrigation water is very
large from May to August. To meet that, snowmelt is stored in dam reservoirs and then used in June
and July, when river discharges decrease.
It is predicted that, as shown by the dotted line in figure 1 the snowmelt runoff will end earlier
in the future due to climate change. In such case, the period when the demand for irrigation water
cannot be met only by river discharge, that is, the time when the water stored in the reservoir is started
to be used, will arrive earlier than at present. This also means a higher probability that the stored water
will become insufficient during the second half of the irrigation period.
The degree to which the snow-melting season will arrive earlier probably varies by the elevation
of the basin. In the case where there are multiple dams and headworks in a single basin but the
catchment of each facility is at a different elevation, the severity of shortfall will vary among those
facilities. In such case, those facilities may have to cooperate in managing water.
In light of this, in the current study, runoff analysis was conducted in a certain area of Hokkaido
with the aim of evaluating the influence of basin elevation on how much earlier the snowmelt starts.
Snowmelt runoff starts Warming accelerates decrease in

earlier due to warming. the amount of stored water.
Reservoir volume
Storage volume
Storage in a reservoir might not be
sufficient due to warming.
River discharge
Demand for paddy-field

irrigation water
Mar. Apr. May Jun. Jul. Aug. Sep.
Figure 1 Conceptual diagram of how warming affects paddy field irrigation in Hokkaido
2. Method
Figure 2 gives an overview of the basin under study. There are dams in two places and
headworks in one place in the basin that were constructed for paddy field irrigation. Figure 3 plots the
elevation distribution of catchments containing the studied facility points. The locations in order of
highest to lowest average elevation are Dam A, Head works C and Dam B.
2 PAWEES 2013 (12TH) INTERNATIONAL CONFERENCE ON AGRICULTURAL WATER AND RURAL ENVIRONMENT FOR THE FUTURE
700～800
Dam A (area weighted
600～700 average elevation : 280m）
500～600 Headworks C (area weighted

Basin of Dam A
Elevation(m)
average elevation : 261m ）
(62.6km2) 400～500 Dam B (area weighted
average elevation : 222m）
Basin of Dam B 300～400
(12.4km2)
200～300
100～200
0～100
Basin of Headworks C 0 10 20 30 40 50 60
(80.7km2) Proportion of area(%)
Figure 2 Inlet works and the size of the Figure 3 Elevation distribution of the basin area
basin containing the studied facility points
In the runoff analysis, as shown in figure 4, a straight 4-stage storage tank model that is suitable
for analyzing long-term runoff was used. The following is the method of formulating the runoff model.
The coefficients of the tank model were determined by referring to recent data, that is, the data on
temperature and precipitation from 1980 to 1999 observed at the AMeDAS observation station nearest
to the basin and the values of runoff volume recorded at each facility location. However, the values of
temperature and precipitation were corrected considering the differences in elevation of the facilities.
The correction method will be explained in the next paragraph. The future runoff was predicted by
entering future temperature and precipitation into the tank model under the assumption that the
coefficients of the tank model created in the above-mentioned method will remain valid in the future.
Note that the discharge at the Head works C was calculated based on the assumption that water storage
was not performed in the upstream dams.
H1
a1(z1)
a2(z2)
b1
H2
a3(z3)
b2
H3
a4(z4)
b3
H4
a5(z5)
b4
Figure 4 Straight 4-stage storage tank model
The values of temperature and precipitation were corrected as follows. First, the basin was
divided by elevation at intervals of 100 m. The predicted temperature and precipitation at the three
study points were obtained by correcting the values of temperature and precipitation recorded at the
observation points in the vicinity according to the elevations of the three points, in which those
measured values were used as basic values. The measured temperature was corrected on the
assumption that temperature would drop by 0.6 degrees per 100 m. The measured precipitation was
corrected, as shown in figure 5, by multiplying the value of daily precipitation observed at the nearest
AMeDAS observations station by 2 coefficients. The first coefficient, which is the ratio between the
“sum of inflow into the dam and evapotranspiration in the basin” and the “total of precipitation
observed at the nearest AMeDAS observations station,” was obtained in order to adjust the water
budget balance during the warm season. The second coefficient is the ratio between the “increase in
precipitation per 100 m in elevation” in the cold season and that in the warm season.
Temperature ：-0.6℃/100m(elevation)
Corrected precipitation value＝daily precipitation observed at the nearest AMeDAS observation station ×1.6×1.5
700
Increase in precipitation per 100m elevation (cold period)
Coefficient ＝
evapotranspiration in the basin（mm）
600 Increase in precipitation per 100m elevation (warm period)

Sum of inflow into the dam and the
500
400
300 Ratio of the two
200
100
0
0 100 200 300 400 500 600 700
Cumulative precipitation observed at the nearest
AMeDAS observation station （mm）
Figure 5 Method of correcting precipitation values
Table 1 The 9 climate models used for analysis
Model （country)
BCCR-BCM2.0 (Norway)
CCSM3 (USA)
CSIRO-Mk3.0 (Australia)
ECHO-G (Germany/ S. Korea)
GISS-EH (USA)
INGV-SXG (Italy)
MIROC3.2 (hires) (Japan)
MIROC3.2 (medres) (Japan)
UKMO-HadGEM1 (UK)
Before entering the values of precipitation into the tank model, it is necessary to determine the
relative contribution of snowfall versus rainfall and to determine whether snowmelt has taken place. In
this analysis, the determination was made as follows: When the daily mean temperature falls below
0 °C in any part of the basin, which is divided according to the elevation, the precipitation is regarded
as snowfall. In addition, in the case that daily mean temperature exceeds 0 °C, the volume of
meltwater was calculated according to the “degree-day method.”
Future runoff volume was predicted by entering the predicted values of temperature and
precipitation of the future period, from 2046 to 2065, into the tank model. For the future values of
temperature and precipitation, the values from the 9 climate models shown in table 1 were used. To
prepare the data for input, “climate change offset” was utilized. This method will be discussed in the
next paragraph. For the greenhouse gas emission scenario, A1B was selected, under which the
concentration of greenhouse gasses in the atmosphere will double in 100 years.
To prepare the data for input, “climate change offset,” which was established by Michihiro et al.
(2011), was utilized. On their home page “Climate Change Information Database for Hydrological
Analysis,” (Water Resources Research Center, 2010) the difference between the present and the future
temperature and precipitation are shown for 29 climate models. For temperature, the change is
expressed in terms of temperature difference; for precipitation, the change is expressed in terms of a
percent increase or decrease. The predicted value of future temperature in a certain area is obtained by
adding the change in temperature to the value from the recent temperature data; the predicted value of
future precipitation in a certain area is obtained by multiplying the value from the recent precipitation
data by a percent increase or decrease. Values of this future change are shown in each 80-km × 80-km
grid cell.
Dam A Headworks C
Black bold line：Average of present situation (1980-1999) Black bold line：Average of present situation (1980-1999)
Other lines：Future forecast of each climate model Other lines：Future forecast of each climate model
Dam B
Black bold line：Average of present situation (1980-1999)

Other lines：Future forecast of each climate model
Figure 6 Results of runoff analysis
3. Results and consideration
Figure 6 shows the prediction results of future runoff at Dam A, Dam B and Head works C. For
each model, the analysis results were obtained by averaging out the values of 15-day moving average
of discharge for the 20-year period. The bold line in the figure represents the hydrograph of current
runoff. At all three points, snowmelt runoff starts earlier than at present, for all the climate models. In
addition, for most of the climate models, the earlier arrival of the peak day of snowmelt runoff and of
the reduction in the peak runoff were observed. As mentioned above, changes were seen in the future
snowmelt runoff in all climate models.
Figure 7 shows the change in snowmelt runoff peak day. The graphs are arranged from top to
bottom in descending order of the average elevation of the catchments. In three basins, the peak day
arrived earlier, but not equally earlier. The higher was the average elevation of the catchment, the
earlier was the peak day, which indicates that the magnitude of the effects of warming varies
depending on the average elevation of the catchment. It is necessary to evaluate the effectiveness of
irrigation water resource management that is cooperatively conducted by water facilities in a close
proximity.
Value of present state obtained

Average of predicted value from AMeDAS data
and standard deviation
Predicted value obtained
Dam A from the model
Headworks C
Dam B
3/27 4/1 4/6 4/11 4/16 4/21 4/26
Figure 7 Change in peak runoff day
Value of present state obtained

from AMeDAS data Predicted value obtained
Average of predicted value
and standard deviation from the model
Dam A
Headworks C
Dam B
80 90 100 110 %
Figure 8 Change in cumulative volume of runoff in the snowmelt season (February-May)
Next, the predicted results of the effect that warming will have on the total runoff will be
discussed. Figure 8 shows the change in the total runoff during the snowmelt season (February-May),
and figure 9 shows the change in the total runoff during the irrigation period (May-August). In these
figures, future changes are expressed as percentages relative to the current total runoff (100%). It is
difficult to make a definite conclusion with respect to how the elevation of the catchment affects the
degree of reduction in these two types of total volume of runoff. However, in most of the prediction
results, the total runoff in the future is lower than that at present. The possible factor that accounts for
the reduction in the total amount of runoff shown in the figures is the reduction in precipitation.
However, the average values of monthly precipitation in the 9 models were close to the values of the
present monthly precipitation. Therefore, it is reasonable to attribute the above-mentioned reductions
in runoff as being due to the decrease in water remaining as snow in the catchments and to the earlier
arrival of snowmelt runoff. The abovementioned decrease can take place in winter in the future when
increases in snowfall and decreases in rainfall are observed, or when increases in snowmelt are
observed
In the hydrographs of predicted results shown in figure 6, the runoff is greater in early winter,
that is, in November and December, than that in the present winter, which indicates that the rainfall as
a share of all precipitation will increase, or the snowmelt amount will be greater in early winter in the
future than it is at present.
Predicted value obtained

Average of predicted value Value of present state obtained
from the model
and standard deviation from AMeDAS data
Dam A
Headworks C
Dam B
50 60 70 80 90 100 110 %
Figure 9 Change in the total volume of runoff in irrigation period (May-August)
4. Conclusion
We conducted runoff analysis in a certain basin in Hokkaido where snowmelt water plays an
important role for irrigation water by using the predicted values of temperature and precipitation for 9
climate models, toward examining the influence of climate change on snowmelt runoff in the future. It
was predicted that the lower is the average elevation of a basin, the earlier is the peak day of snowmelt
runoff and the lower the runoff will become in the snowmelt season and in the irrigation period. In the
future, we will quantitatively show the change in snow water equivalent that is stored in the basin and
examine the mechanism of how warming affects snowmelt runoff.
References
MICHIHIRO, Y., SUZUKI, Y., and SATO, Y., 2011, “Development of Climate Change Information
Database.” Journal of JSCE, Division B2, 67(2), 1211-1215. (in Japanese)
Water Resources Research Center, Disaster Prevention Research Institute, Kyoto University, 2010
Climate Change Information Database for Hydrological Analysis,
http://hes.dpri.kyoto-u.ac.jp/database/index.html?LANG=EN
WATER AND RURAL ENVIRONMENT
FOR THE FUTURE
A-03
Assessment of Climate Change Impact on Crop Yields in

Northern Taiwan by Using Principal Component Analysis
RAY-SHYAN WU i*, MING-HSU LI**, JI-TANG FANG***, CHI-MEI WANG****

*Professor 1Corresponding author: Dep. of Civil Engineering , National Central University,
No.300, Jhongda Rd., Jhongli City, Taoyuan County 32001, , Taiwan (R.O.C.), TEL:
+88634227151 ext.34126, raywu@ncu.edu.tw.
**Associate Professor and Director, Institute of Hydrological and Oceanic Sciences, National
Central University.
***Master, Dep. of Civil Engineering, National Central University.
****PhD, Dep. of Civil Engineering, National Central University.
ABSTRACT
Climate change affects all agriculture activities. When long-term climate pattern has changed, the
weather factors, such as temperature and rainfall, might affect the quality and quantity of crop
growth. Paddy rice is the most important crop productions in Taiwan and accounts for more than 70%
of total water resources usage. The quantity of rice productions is a very important index for food
security and agriculture management. This study utilizes the Decision Support System for
Agrotechnology Transfer (DSSAT) model to analyze the variations of growth days and quantity of
paddy rice under climate change. The Weather Generator Model (WGEN) was used to generate daily
rainfall and daily mean temperature. Maximum and minimum daily temperature and solar radiation
were then estimated by regression functions of daily mean temperature and daily rainfall from
historical data. Rice productions were estimated by the DSSAT model. The Taiwan Climate Change
Projection and Information Platform Project (TCCIP) provides future climate projections and the
A1B scenarios of Special Report on Emissions Scenarios were selected in this study. To understand
the dominant factors affecting crop yields under climate change, the Principal Component
Analysis(PCA) was applied to analyze DSSAT results for both periods of baseline data (1985~1990)
and near future data (2020~2039). Accumulated solar radiation, accumulated growing degree, crop
water requirement and growing days were retrieved for performing PCA. Climate variations
projected by ensemble models and CM3 model showed accumulated growing degree before blossom
is the most important factor, while in MK3_0 mode is the accumulated solar radiation before
blossom.
Keywords: Climate change, rice, DSSAT, temperature, water requirement
1. Introduction
1.1 Motive
Climate change causes temperature and rainfall pattern changes. Weather condition changes
affect the crop growth environment, growth period, and seasonal patterns, as well as crop quality and
quantity. Chen (2012) considered that the natural resources required for growing crops, such as water,
soil, and gene heredity, decrease or deteriorate when exposed to the detrimental effects of global
warming, uncertain rainfall, and extreme weather. A decrease in the crop quality and quantity
damages food security.
The climate greatly affects agricultural activities. When the climate changes, it also affects
water demand and crop quantity. In Taiwan, the water demand for agriculture is approximately 70%
of the total yearly water demand. The water demand for the irrigation of crops is approximately 60%
of the total water demand for agriculture, which is the largest amount of agricultural water usage.
Because water for agricultural use has a better lack of water tolerance than water for domestic and
industrial use, it is frequently transferred to other supply sources during droughts. Therefore, when
irrigation water usage changed, the difference in the water usage distribution might affect crop yields.
The interaction of crop growth and the meteorological environment is extremely complex. Various
actions and reactions occur at different times and places. To estimate the effect of crop growth on
climate change, a crop model is an appropriate tool to use for analysis and discussion. A crop model
requires various climate, hydrologic, and physiographic input data. Certain climate data, such as
temperature, rainfall, and solar radiation, are widely used to discuss the influence of crop yield
variation.
In this study, the Decision Support System for Agrotechnology Transfer (DSSAT) crop model
(Hoogenboom et al., 2010) was used to simulate the crop yields under different with climate scenarios.
To estimate future crop variation, simulations of the main climate input data of the model, daily
maximum and minimum temperature, daily rainfall, and daily radiation were necessary. These future
data were generated using a Weather Generation Model (WEGM) and by consulting official public
literature. The crop yield is influenced by multiple variables, such as climate data, crop genetic
information, and soil data. To simplify the analysis, climate data (temperature and radiation) were
chosen as the variables in this study. The relationship between the multiple variables was not
independent. To analyze the complex crop yield data, Principal Component Analysis (PCA) a method
of multivariable statistical analysis, was used to understand the critical components of yield influence.
The analysis result can be an index for the establishment of future food security.
1.2 Literature review

Regarding global warming, the Intergovernmental Panel on Climate Change (IPCC) published
its fourth assessment report in 2007. It contained an analysis of the surface temperatures of various
areas worldwide. The results showed that the global average temperature continues to increase. The
linear trend was 0.74 °C ± 0.18 °C in the last century (1906-2005). The range of temperature increase
in the last 50 years (1956-2005) was double that of the first half of the last century.
Paddy rice is the most crucial crop in Taiwan. Its yield is also a key index of food security and
agricultural cultivation. Climate change causes temperature and rainfall pattern changes, which in
turn cause crop yield and growth changes. Therefore, climate change is a factor that greatly influences
crop yield and growth.
Crop models are highly useful for crop assessment because the interactions between crops and
the climate during the growth process are complex. Many crop models include climate assessment
ability for estimating crop growth in different times and areas. The DSSAT model has been widely
used in rice-cultivating countries for estimating the impact of climate change (Tsvetsinskaya et al.,
2003; Mall & Aggarwal, 2002). Zhou (2004) used the DSSAT model to estimate the rice yield under
simulated future Taiwanese climate conditions. When the cultivation schedule was fixed and the
future concentration of carbon dioxide was not considered, the results showed that the crop yield
decreased in both the first and second rice growing periods. Jansen (1990) used the Modules of an
Annual Crop Simulator (MACROS) model to investigate rice growth under climate-change-affected
conditions in South East Asia. The results showed that the yield increased when the temperature was
slightly increased. However, if the temperature was increased to over 0.8 °C, the yield decreased. Yao
and Chen (2009) investigated the difference of rice growth and yields for different climate patterns by
using the DSSAT model. They also discussed how three climate factors (temperature, day radiation,
and rainfall) influenced rice yields. The results showed that the DSSAT model was suitable for use in
Taiwan. The yield observation data and the DSSAT simulation data exhibited strong correlation, and
the correlation coefficient was between 0.7 and 0.8.
A thesis by Yang (2007) reported that the greenhouse effect hastens global warming and
climate change. Consequently, not only do irregular weather patterns occur frequently, these
irregularities also cause instability and uncertainty regarding water resource levels. To improve the
efficiency of agricultural water usage, irrigation techniques and methods should be reviewed.
Actively increasing water usage efficiency might provide the largest output benefit in locations with a
limited water supply. Su (2009) summarized several studies regarding global-warming-influenced
crop production and observed that global warming has shortened the crop growth period and
decreased crop production in Taiwan. It was shown that increased temperature at night positively
influenced rice yield and quality.
Hsu (2009) analyzed the long-term time and space hydrology observation data variability in the
Tamsui river basin. The data years were between 1961 and 2007. Linear regression and the
nonparametric tendency test were used to estimate the rate of time series change. PCA was used to
conduct a space variability analysis. Zhang (2009) discussed the change of water quality factors for
location and season. A multivariate statistical analysis model was used to analyze the factors of water
quality. PCA can be used to analyze multiple water quality-influencing components and explain the
essential factors.
2. Model Introduction
2.1 DSSAT
DSSAT is a crop growth simulation system developed by the International Benchmark Sites
Network for Agrotechnology Transfer (IBSNAT) to investigate the influence of the environment and
weather on crops. DSSAT and its crop simulation models have been used for many applications
ranging from on-farm and precision management to regional assessments of the climate variability
and climate change impact. It has been in use for more than 20 years by researchers, educators,
consultants, extension agents, growers, and policy decision makers in over 100 countries worldwide
(Jones et al., 2003).
The system combines crop, soil, and weather data and applies the data to different programs,
which users can manipulate to simulate growing different crops in varying conditions worldwide. The
simulation results can then be used to quickly estimate the new crop cultivation method.
Every rice growing period that the DSSAT model calculated depended on the growing degree
days (GDD) method. The basic definition of a growing degree is the sum of the deviation of the daily
maximum and minimum temperature average minus the base temperature of a particular growing
period (eq. 1). The GDD is the degree of growth; Tmax is daily maximum temperature (°C); Tmin is
daily minimum temperature (°C); and b is base temperature, the temperature at which crops stop
growing. For example, the base temperature of rice was 9 °C (Yao et al., 2000). If GDD<0, GDD
equals 0.
GDD = ∑((Tmax + Tmin )/2 − b) (1)
The input data of the DSSAT model included four main parts: weather, soil, crop genetic
information, and management. According to the DSSAT 4.0 user’s manual, the minimum required
weather data must include daily maximum and minimum temperature, daily rainfall, and daily
radiation. The length of weather data must cover the entire growing process. The soil data must
describe the soil’s characteristics, such as soil drainage, runoff, evaporation, radiation reflection
factor, water content in different layers, upper and lower limits of soil water content, and soil nitrogen
and phosphorous content. The management strategy required plant density (including seeding and
transplantation), plant depth, planting date, latitude of plant area, and irrigation management method
(including irrigation method, irrigation water demand, and irrigation timing). The crop growth period
and biomass accumulation during the growing period were decided using crop genetic information.
The rice variety used in this study is called Tainung No. 67, the most cultivated variety in Taiwan.
The official version of Tainung No. 67’s crop genetic information was used (Yao et al., 2000) as the
input data.
2.2 Weather generator model

Pickering et al. (1988) and Richardson (1983) developed the Weather Generator Model
(WGEN). Tong (2003) used it as a base to generate future temperature and rainfall data. The daily
temperature with future weather conditions used the monthly average temperature and was simulated
�j was the jth month’s
using the first order Markov chain (eq. 2). Ti was the ith day’s temperature; T
average temperature; ρ was the first order serial correlation coefficient of Ti and Ti−1 in the jth
j
month; Ni was a random number between 0 and 1; and σ was the standard deviation of historical
j
data in the jth month.
2
�j + ρ �Ti−1 − T
Ti = T �j � + Ni σ �1 − ρ (2)
j j j
It was assumed that the first day’s temperature was replaced with the monthly average
temperature at present month. Using the statistical characteristics of historical monthly average
temperatures, the daily temperature could be generated.
The daily rainfall simulation was divided into two parts–rainfall event differentiation and
rainfall amount estimation. The simulation of rainfall event differentiation depended on historical
rainfall probability. In every month, P(W|W) was the probability of both rainfall on the Ith and I-1th
day; P(W|D) was the probability of rainfall on the Ith and not on the I-1th day. The generator produced
a random number (RN) between 0–1. Every first day of the month, when the RN was less than or
equal to the rainfall probability P(W), it was a rainy day. Excluding the first day of every month, the
rainfall conditions of the previous day were used to determine that the whether the following day
would be a rainy. According to the historical data probability of P(W|W) or P(W|D), if the RN was
less than or equal to P(W|W) or P(W|D), it would be a rainy day. The Ith day rainfall event
discriminant is as follows:
a. If the rainfall number at the I-1th day was larger than 0, and when RN ≤ P(W/W), the Ith day
was a rainy day. Otherwise, it was not.
b. If the rainfall at the I-1th day equaled 0, and when RN ≤ P(W/D), the Ith day was a rainy day.
Otherwise, it was not.
Hong (2007) used exponential distribution to simulate ideal daily rainfall. The exponential
distribution equation is shown in eq. 3. The P was daily rainfall (cm/day), P �ȷ was rainy day average
rainfall at the jth month (cm), and N was a random number between 0 and 1.
P = P�ȷ × (− ln(1 − N) (3)
2.3 Irrigation demand estimation

To determine the irrigation requirement, the main factor was crop evapotranspiration. Crop
evapotranspiration (ETi) could use potential evapotranspiration (ET0i) and the crop coefficient (Kc) to
estimate (eq. 4). The Ritchie modification of the Priestley-Taylor method (Priestly & Taylor, 1972;
Ritchie, 1972, 1985) was used to calculate the potential evapotranspiration in this study.
ETi = ET0i × K c (4)
2.4 The theory of Principal Component Analysis (PCA)

PCA is a multivariate statistical method. It was provided by Karl Pearson in 1901 and
developed by Hotelling in 1933. The main purpose of this analysis method was to use reduced data
and a simplified structure. Using only a few principal components could replace multi-original
explanation variables. The principal components transferred to a few independent linear combinations
and generalized to the integer index (Fu, 2002).
If there were m original variables and n observation samples, they could construct an nm
original data matrix P (eq. 5).
x11 x12 … x1m

x21 x22 … x2m
P=� ⋮ ⋮ ⋱ ⋮ � (5)
xn1 xn2 … xnm
Because the unit of original data was different, to reduce the calculation deviation, the data
must be normalized to made average was 0 and the variance was 1. After normalization, it could
calculate the eigenvalues (λ ) and eigenvectors of the matrix. The principal component was the new
i
variable to which the eigenvector corresponded. The new variable variance was the eigenvalue. The
new variance explained the variance ratio with the percentage of the total eigenvalue.
The eigenvector was the component loading. The relation factor between the new and old
variables showed the influence and importance of the old variable to the new variable. If the value
was larger than 0.75, the relation was high; between 0.5 to 0.75, the relation was normal; between 0.3
to 0.5, the relation was low (Liu et al., 2003; Zhang, 2009). The proportion of the component loading
was used to explain the meaning of the component (eq. 6).
𝒲ij �λ
i
ℒij = = ryi xj (6)
𝒮j
The variable ℒij was the jth variable on the ith component loading; 𝒲ij was the jth variable
on the ith component weight; λ was the ith eigenvalue of the component; 𝒮j was the standard
i
deviation of the jth variable. Several methods were used to choose components: (a) The percentage of
variance accumulation, (b) rule of thumb, (c) the scree test, and (d) deletion of the component with
low explanation ability.
3. Climate change scenario
The default scenarios for future climate analysis presented in the Special Report on Emission
Scenarios (SRES) provided by the IPCC was adopted to simulate and analyze climate change. The
default scenarios in SRES belongs to the greenhouse gas emission scenario, which proposed four
main possible scenarios of greenhouse gas emission (A1, A2, B1, and B2) from an economical,
population, industrial, and environmental perspective. The A1B scenario was used in this study. With
the balanced development of every energy resource, it was progressive and the most likely future
scenario.
The future climate scenarios of climate change research involve assumption, time or space
analogy, and simulations using General Circulation Models (GCMs). Global Climate Models (GCMs)
used for climate studies and climate projections were conducted at a coarse spatial global resolution
and were unable to resolve sub-grid scale features in particular areas. As a result, GCM output could
not be used for local impact studies before the downscaling process. In this study, downscaling data
obtained from the Taiwan Climate Change Projection and Information Platform Project (TCCIP)
were applied to three GCM models: CNRM_CM3, CSIRO_MK3_0, and 24 sets of GCM downscale
ensemble models to downscale from a global scale to a 25 × 25 km watershed scale, under which the
simulation of hydrological change in Taiwan was for the near future (2020-2039) on the baseline of
1985 to 1999 (table 1). The GCM grid (lat.: 24.88, long.: 121.13) used was close to the Taoyuan
District Agricultural Research and Extension Station (lat.: 24.95, long.: 121.03).
Table 1. Difference of temperature and rainfall by GCM models downscaling scenario data
Temperature (°C) Rainfall (%)

CSIRO_MK3_ CSIRO_MK3_
month CNRM_CM3 Ensemble CNRM_CM3 Ensemble
0 0
1 0.79 1.12 0.93 14.67 21.66 -0.24
2 0.40 0.94 0.82 -9.82 67.08 -0.14
3 0.32 0.85 0.75 -30.82 38.08 -6.75
4 0.31 0.45 0.68 -6.18 9.44 -4.77
5 0.73 0.19 0.74 -23.00 -1.46 1.24
6 0.90 0.38 0.65 -10.02 -15.46 1.18
7 1.00 0.40 0.68 -7.32 46.16 10.60
8 0.84 0.26 0.69 3.60 8.36 2.78
9 0.37 0.57 0.71 -3.17 7.50 5.18
10 0.31 0.43 0.62 -31.76 1.37 -5.02
11 0.66 0.73 0.77 0.27 0.38 -6.22
12 0.33 0.89 0.90 -10.99 22.71 1.82
If the climate change in the study area equaled the prediction of the nearby GCM grid, the
equations could be modified as eq. 7 and eq. 8. The m was 1 to 12 months; CTm was the daily average
temperature in the jth month after modification; CPm was the rainy day average rainfall in the mth
month after modification; Tm was the difference of temperature in the GCM model scenario; RPm
was the rainfall ratio in the GCM model scenario. The modification results were used to generate the
future average temperature and rainfall, using WGEN.
CTm = Tm + ∆Tm (7)
CPm = Pm × RPm (8)
4. Result and discussion
4.1 Crop yield simulation using the DSSAT model

This study simulated the first period of rice crop yield in the Taoyuan area of Taiwan. To
confirm that the DSSAT model could respond well to the crop yield of the study area, the real crop
yield in 2008 was used to test the data, and the real crop yield in 2007 was used to model the
verification data. The irrigation in the model was set to have no lack of water. The results showed that
the deviation between the real values and simulated values was less than 5% (table 2).
Table 2. The test and verification of DSSAT model
Simulation yield (DSSAT)

year Real yield 1) (kg/ha) Deviation (%)
(kg/ha)
2007 5288 5441 3.77%
2008 5416 5620 2.89%

Note) 1) The Council of Agriculture, 2008.
Considering no carbon dioxide influence (concentration was 360 ppm), the yield change rate of
the three model (Ensemble model, CM3 model and MK3_0 model) simulations on the A1B scenario
was between -2.3% and -5.6% (figure 1). All the future crop yields decreased when the future
temperature increased. These results were the same as the conclusion of Yao (2009).
系集
Ensemble CM3 MK3_0
0
-1
Yield Change(%)
-2
-3
-4
-5
-6
Figure 1. The change of crop yield on each GCM model
4.2 Principal Component Analysis

In this study, weather data from the baseline and the three scenario models were used to
generate 100 future weather data sets of each model by using the WGEN model. Each of the 100 sets
of weather generation data were used to simulate the first period crop yield. The crop yield variables
could then be chosen for further analysis, which included eight factors: the accumulated solar
radiation before and after blossoming (SRAD_B / SRAD_A), accumulated crop water requirement
before and after blossoming (WATER_B / WATER_A), accumulated degree of growth before and
after blossoming (TEMP_B / TEMP_A), and accumulated days of growth before and after
blossoming (DAY_B / DAY_A).
The Kaiser-Meyer-Olkin (KMO) test and Bartlett Sphericity test were used to determine
whether the variables were adequate for PCA. When the value of KMO sampling adequacy was
larger than 0.5, it proved the meaning of the original variable. When the value of the Bartlett test was
less than 0.05, the variable was significant. The results showed that all four cases passed the tests
(table 3).
Table 3. KMO and Bartlett test
Model Test Value

Kaiser-Meyer-Olkin MSA 0.506
Baseline
Bartlett Sphericity Test Significance 0.000
Ensemble
CM_3
MK3_0
After confirming the adequacy of the variables, the eigenvalue of the correlation coefficient
matrix and the explanation of variance were calculated, and the component loading matrix was
eventually obtained. There were eight principle components in this study. However, for the purpose
of dimension simplification, the component analysis only included eigenvalues larger than 1. The
results showed that the baseline had three components and that the accumulated explanation of
variance was 74.53%; the other models all had four components and the accumulated explanations of
variance were 96.93%, 97.67%, and 98.95% (table 4). The component loading matrix of each model
is shown in table 5.
Table 4. Eigenvalue and explanation of variance of models
Principle Explanation of variance Accumulation of explanation

Model Eigenvalue
component (%) variance (%)
1 2.60 32.40 32.40
2 2.01 26.22 58.62
3 1.27 15.91 74.53
4 0.92 11.50 86.03
Baseline
5 0.60 7.41 93.44
6 0.40 5.02 98.46
7 0.11 1.42 99.88
8 0.01 0.12 100.00
1 3.19 39.90 39.90
2 2.05 25.65 65.55
3 1.46 18.31 83.86
4 1.05 13.07 96.93
Ensemble
5 0.19 2.36 99.29
6 0.03 0.39 99.68
7 0.02 0.26 99.94
8 0.01 0.06 100.00
1 3.23 40.42 40.42
CM3 2 2.11 26.42 66.84
3 1.38 17.26 84.10
4 1.09 13.57 97.67
5 0.15 1.81 99.48
6 0.03 0.37 99.85
7 0.008 0.10 99.95
8 0.004 0.05 100.00
1 3.07 38.33 38.33
2 2.29 28.68 67.01
3 1.44 17.96 84.97
4 1.12 13.98 98.95
MK3_0
5 0.05 0.59 99.54
6 0.02 0.22 99.76
7 0.01 0.14 99.90
8 0.008 0.10 100.00
The explanation of variance of the first component of the baseline model was 32.4%, which
signifies that the accumulated degree of growth before and after blossoming and the accumulated
days of growth before and after blossoming were highly correlated. Those four components were also
the most critical crop yield-influencing factors in the Taoyuan area. Because the accumulated degree
of growth after blossoming was higher than that before blossoming, the first component could be the
temperature after blossoming. The explanation of variance of the second component was 26.22%,
which signifies that the accumulated degree of growth before blossoming, the accumulated days of
growth after blossoming, and the accumulated crop water requirement after blossoming were highly
correlated. Hence, the second component could be the temperature before blossoming. The
explanation of variance of the third component was 15.91%, which correlates the accumulated
requirement before blossoming with the accumulated solar radiation before blossoming. The results
indicated that the higher the solar radiation, the higher the crop water requirement. Hence, the third
component could be solar radiation. According to the results of the component analysis, the main
component sequences that influence crop yield are the temperature after blossoming, the temperature
before blossoming, and solar radiation. The summarized results of all the models are shown in table 6.
Table 5. The component loading matrix of models

(a) Baseline
Component
Crop yield variable
1 2 3 4
SRAD_B -0.45 0.33 -0.54 ---

WATER_B 0.20 0.16 0.80 ---
TEMP_B -0.62 0.74 0.18 ---
DAY_B -0.62 0.71 0.23 ---
SRAD_A 0.48 0.70 -0.13 ---
WATER_A 0.15 -0.14 0.47 ---
TEMP_A 0.80 0.51 -0.10 ---
DAY_A 0.81 0.36 -0.11 ---
(b) Ensemble
Component
Crop yield variable
1 2 3 4
SRAD_B 0.81 -0.26 0.30 0.41

WATER_B 0.69 -0.41 0.19 0.56
TEMP_B 0.82 0.16 0.33 -0.43
DAY_B 0.79 0.19 0.32 -0.48
SRAD_A 0.27 0.90 -0.18 0.27
WATER_A 0.38 0.84 -0.34 0.18
TEMP_A -0.49 0.43 0.67 0.22
DAY_A -0.54 0.27 0.73 0.04
(c) CM3
Component
Crop yield variable
1 2 3 4
SRAD_B 0.87 0.08 0.11 0.46

WATER_B 0.76 -0.01 -0.03 0.65
TEMP_B 0.84 0.21 0.35 -0.35
DAY_B 0.81 0.21 0.36 -0.41
SRAD_A -0.01 0.79 -0.55 0.14
WATER_A 0.26 0.76 -0.45 -0.28
TEMP_A -0.47 0.76 0.38 0.20
DAY_A -0.52 0.48 0.69 0.16
(d) MK3_0
Component
Crop yield variable
1 2 3 4
SRAD_B 0.84 -0.01 0.26 0.45

WATER_B 0.75 -0.09 0.18 0.62
TEMP_B 0.83 0.09 0.28 -0.48
DAY_B 0.80 0.10 0.21 -0.54
SRAD_A 0.25 0.86 -0.43 0.07
WATER_A 0.32 0.73 -0.59 0.03
TEMP_A -0.33 0.81 0.47 0.05
DAY_A -0.45 0.58 0.67 0.05
Table 6. The PCA result of models
Component
Model
1st 2nd 3rd 4th
Baseline TEMP_A TEMP_B SRAD ---

Ensemble TEMP_B SRAD TEMP_A WATER_B
CM3 SRAD_B SRAD_A DAY WATER
MK3_0 SRAD_B SRAD_A DAY WATER
5. Conclusion
In this study, the DSSAT model was used to simulate the crop yields with baseline and three
different GCM models’ weather data on the A1B scenario. The results showed that the DSSAT model
could well respond to the crop yield of the study area. All the future crop yields of first period
decreased when the future temperature increased.
The climate greatly affects agricultural activities. To analyze the complex crop yield data of
climate change, PCA was used to understand the critical components of yield influence. The results
showed that the baseline model had three components and that the accumulated explanation of
variance was 74.53%; the other models all had four components and the accumulated explanations of
variance were 96.93%, 97.67%, and 98.95%. The most important factor of baseline model was the
accumulated degree of growth after blossoming; the ensemble model was the accumulated degree of
growth before blossoming; the other models were the accumulated solar radiation before blossoming.
The results indicated that the main cause of yield influence was the accumulated degree of growth
after blossoming at present. Under the climate change, the main cause of yield influence might
transfer into the accumulated degree of growth before blossoming or the accumulated solar radiation
before blossoming.
Reference
Chen, T. Y., 2012, Impact of Climate Change on Paddy Rice Yields and Irrigation Water Requirement
in Taoyuan, Master thesis, National Central University.
Fu, C. X., 2002, “Issues Related to Principal Component Analysis and Common Factor Analysis.”
Formosan Education and Society, Vol.3, 107-132.
Hong, N. M., 1996, The Impacts of Climate Change on the Management of Water Resources in the
Ta-An River, Master thesis, National Taiwan University.
Hoogenboom, G., J.W. Jones, P.W. Wilkens, C.H. Porter, K.J. Boote, L.A. Hunt, U. Singh, J.L. Lizaso,
J.W. White, O. Uryasev, F.S. Royce, R. Ogoshi, A.J. Gijsman, G.Y. Tsuji, and J. Koo., 2010,
Decision Support System for Agrotechnology Transfer (DSSAT) Version 4.x [CD-ROM].
University of Hawaii, Honolulu, Hawaii.
IPCC., 2007, “Climate Change 2007：The Physical Science Basis. The AR4 Synthesis Report.”
Jansen, D.M., 1990, “Potential rice yields in future weather conditions in different parts of Asia.”
Netherlands Journal of Agricultural Science, 38(4), 661-680.
Jones, J.W., G. Hoogenboom, C.H. Porter, K.J. Boote, W.D. Batchelor, L.A. Hunt, P.W. Wilkens, U.
Singh, A.J. Gijsman, and J.T. Ritchie., 2003, “DSSAT Cropping System Model.” European
Journal of Agronomy,18, 235-265.
Liu, C. W., Lin, K. H, Kuo, Y. M., 2003, “Application of factor analysis in the assessment of
groundwater quality in a Blackfoot disease area in Taiwan.” The Science of the Total
Environment, 313, 77-89.
Mall, R. K. and Aggarwal, P. K., 2002, “Climate Change and Rice Yields in Diverse
Agro-Environments of India. I. Evaluation of Impact Assessment Models.” Climate Change, 52,
315-330.
Pickering, N. B., Stedinger, J.R., and Haith, D.A., 1988, “Weather input for nonpoint source pollution
models”, Journal of Irrigation and Drainage Engineering, 114(4), 674-690.
Priestley, C. H. B. and Taylor, R. J., 1972, “On the assessment of surface heat flux and evaporation
using large-scale parameters.” Mon. Weather Rev., 100, 81-82.
Richardson, C. W., Foster, G. R., Wright, D. A., 1983, “Estimation of erosion index from daily
rainfall amount.” Transactions of the ASABE, 26 (1), 153-156.
Ritchie, J. T., 1985, “A guaer-oriented model of the soil water balance in wheat.” Wheat Growth and
Modeling, Plenum Publishing Corporation, USA, 293-305.
Ritchie, J. T., 1972, “Model of predicting evapotranspiration from a row crop with incomplete
cover.”, Water Resources Research, 8(5), 1204-1213.
Su, Z. Z., 2009, “The Response Strategies of Food Production under Climate Change in Taiwan.”
Agriculture Policy & Review, 200, 37-40.
Taiwan Climate Change Projection and Information Platform Project (TCCIP),
2012, http://tccip.ncdr.nat.gov.tw/NCDR/main/index.aspx.
The Council of Agriculture, 2008, Yearbook of Taiwan Agriculture Statistics. Taipei: The Council of
Agriculture, Executive Yuan, R.O.C.
Tong, Q. B., 2003, Climate Change Impacts and Adaptations for National Communication Under the
United Nations Framework Convention on Climate Change – Climate, Hydrology, and Ecology
(II), Environmental Protection Administration, R.O.C. EPA-92-FA11-03-034. Department of
Bioenvironmental Systems Engineering, National Taiwan University.
Tsvetsinskaya, E. A., Mearns, L. O., Mavromatis, T., Gao, W., Mcdaniel, L., and Downton, M. W.,
2003, “The Effect of Spatial Scale of Climate Change Scenarios on Simulated Maize, Winner
Wheat, and Rice Production in the Southeastern United States.”, Climate Change, 60, 37-71.
Xu, D. M., 2009, Analyses of Temporal and Spatial Hydrological Variations in the Tamsui River
Basin, Master thesis, National Central University.
Yang, C. M., 2007, “The Saving Water Irrigation Cultivation Techniques and Strategies of Paddy Rice
under Global Warming.” Paper presented at the Symposium of Effects of Global Warming on
Rice Industry in Taiwan. Taichung, October, 26.
Yao, M. H., and Chen, S. H., 2009, “The Impact Evaluation of Climate Change on the Growth and
Yield of Rice.” J. of Crop, Environment & Bioinformatics, 6(3), 141-156.
Yao, M. H., Lu, H. S., Zhu, J., and Cai, J. C., 2000, “The Applicability of DSSAT Model to Predict
the Production of Rice and to Evaluate the Impact of Climate Change.” J. of Agriculture
Research of China, 49(4), 16-28.
Zhang, C. E., 2009, Water quality assessment using multivariate statistical techniques in Yuan-Yang
Lake, Master thesis, National United University.
Zhou, M. J., 2004, Influences of Climatic Change and Irrigation Water Transfer on the Potential
Production of Rice, Master thesis, National Taiwan University
THE FUTURE
A-06
Climate Change Action Plan for Water Resources in Taiwan
Wei-Fuu Yang*, Chi-Ming Chen**, Pei-Jung Wu***

* Director-General of Water Resouece Agency, Ministry of Economic Affairs, ROC(Taiwan)
** Manager of Sinotech Engineering Services, ROC(Taiwan)
*** Team leader of Sinotech Engineering Services, ROC(Taiwan)
ABSTRACT
In the paper, the progress of strategies and actions plans in water resources for adaptation under the
climate change are presented. The goal of water resources management in Taiwan is “to ensure the
sustainability of water resources and maintain a balance between the supply of and demand for
water”, which is declared in “Adaptation Strategy to Climate Change in Taiwan” (Council for
Economic Planning and Development, 2012) through support of Taiwan Water Resources Agency
(WRA) and several departments in Executive Yuan. WRA clarified climate change impacts on water
resources in Taiwan, and established the adaptation action plans. In 2009, the historical observation
data was analyzed to discover the climate change trends of precipitation and temperature in Taiwan,
and the three key issues of water resources management were identified as well. In addition, future
scenarios were set based on the projections by global circulation models (GCMs), and the impacts on
water resources management were evaluated. The projections result in the most likely scenario shows
that “the wet season getting wetter and the dry season getting dryer”. The major effects of climate
change are: rising global temperatures, change in precipitation patterns, increasing frequency of
extreme weather phenomena, and raising sea levels. The degree of hazard, vulnerability, and risk of
climate change impact were assessed. The three key risks on water environment, (1) the decreasing of
surface water resource, (2) the decreasing of reservoir storage, and (3) the increasing of agricultural
water demand, are addressed. The adaptation action plans were developed for the high-risk area.
Furthermore, through the following researches in future, the adaptation action plans will be adjusted
periodically to enhance the adaptation capacity of water resources in Taiwan.
Keywords: climate change, water resources, adaptation
1. Introduction
To respond to climate change impact, to strengthen adaptive capacity, to lower social

vulnerability and to establish integrated operating systems, the Council for Economic Planning and
Development (CEPD), Executive Yuan, has drawn up the National Climate Change Adaptation Policy
Framework which serves as the foundation for implementing a comprehensive strategy in the face of
global climate change. There are 8 sectors of the national framework which are Disasters,
Infrastructure, Water Resources, Land Use, Coastal Zones, Energy Supply and Industry, Agricultural
Production and Biodiversity, and Health. Water Resources Agency (WRA) is assigned to be the lead
agency of water resources sector.
As defined in the National Climate Change Adaptation Policy Framework, the objective of
water resources adaptation is to ensure the sustainability of the nation's water resources and maintain a
balance between the supply of and demand for water. To implement water resources adaptation, 4
adaptation strategies and 14 corresponding measures are planned (table 1) (CEPD, 2012).
Referring to world trends and recommendations of the Intergovernmental Panel on Climate
Change (IPCC) (IPCC, 2007), the steps of water resources adaptation planning and promotion in
Taiwan are: (1) Analyzing water resources status and climate change projections; (2) Establishing
adaptation ranges and key issues; (3) Setting hydrological scenarios under climate change situation; (4)
Assessing impacts on water resources; (5) Assessing the risks of water resources; (6) Planning and
promoting actions.
Table 1 Adaptation strategies and corresponding measures in water resources sector
Adaptation Strategies Adaptation Actions
Overseeing all developmental or construction projects. While

analyzing the costs and benefits and evaluating the influence on the
Taking sustainability as
1-1 environment, it is impor tant to also consider climate change and the
the highest guiding
nation's water resources to prevent the current aqua environment from
principle for the nation's
being upset.
1 water resources
Systematically planning and managing the governance of river basins.
management, and also
The planning and management should be based on each river basin's
focusing on protecting
1-2 special characteristics, while taking the aqua environment pollution
the aqua environment.
control, sustainability of freshwater resources, and protection of
biodiversity and ecosystems into consideration.
Revitalizing current water storage capacities. Repairing and
2-1 maintaining related facilities when needed, mostly to prevent
unnecessary loss of water through leakage or during transportation.
Implementing the reservoir watershed land use management, proper
Reviewing water
use of water resource operating funds to promote reservoir
resources management
2-2 watershed conservation. Converting farmland into wo o dlands to
from the supply side and
2 prevent agricultural pollutants like fertilizers and pesticides from
reinforcing the efficient
contaminating reservoir waters.
usage of water
Reinforcing cross-regional conjunction use of surface and ground
resources.
2-3 water. Rewarding the development, promotion, and application of
alternative water resources such as rainwater or reclaimed water.
2-4 Strengthening the contingency measures for unusual water shortages.
2-5 Implementing water rights administration.
Establishing a database Adjusting the water price to a reasonable level, changing water use
3 of the regional total 3-1 patterns, establishing a reasonable, fair, and flexible transformation
quantity of water supply mechanism for water utilization, promoting water conservation efforts,
system and reviewing reexamining the current architecture laws, strengthening the regulation
water resource total of public buildings and facilities, and promoting the installation of
quantity control policies water-conserving devices in public properties.
from the demand side in
Encouraging the development of low waterconsumption industries.
order to reinforce the
3-2 Products that require high water-consumption during production can
efficiency of water
potentially be imported from water-abundant nations.
resources.
Adjusting the agricultural system by considering the environmental
sustainability and farmland productivity. If the two conditions are
3-3 sustained, we can then promote precise irrigation and improve
irrigation methods, which would improve rainwater resource efficiency
and reduce the demand for irrigation.
Requiring products to be printed with water consumption labels for the
4-1 consumers' reference to reduce the consumption of products with high
water usage.
Encouraging corporations to establish waterconserving processes of
4-2
production in order to decrease the water consumption.
Promoting water Creating financial incentives for conserving and recycling water
4-3
sustainability in line resources.
4
with the United Nations' Calculating water accounts through systematic analysis using the
water footprint concept. concept of material flow and water balance. Examining the
reasonableness of the monitoring data of river basins from dif ferent
4-4 government depar tments and gaining a full insight into critical
environmental information such as that on the atmosphere, quality and
quantity of water. Analyzing the water footprint and water resources
utilization in an appropriate manner.
2. Water resources in Taiwan
2.1 Demand and supply

In Taiwan, rainfall is one of the main resources of water supply. The average annual rainfall
amount is about 95.07 billion tons, and the loss from evapotranspiration is about 21% of rainfall (20
billion tons). In the rest parts, annually 70.1 billion tons (74% of rainfall) becomes river run-off, and
4.97 billion tons (5% of rainfall) is groundwater infiltration (WRA, 2011).
The average annual water consumption in Taiwan is about 21 billion tons, and most is from
agricultural consumption, which is about 12.93 billion tons (71%). Domestic consumption is about
3.53 billion tons (20%), and industrial consumption is about 1.66 billion tons (9%) (WRA, 2011).
3 main ways for water supply in Taiwan are: (1) river water diversion, (2) reservoir water
supply and (3) ground water pumping. Reservoirs are the most important infrastructures of water
resources. Supply from them is 24% of total. In recent years, because of increasing frequency of
high-intensity storm events, the sedimentation of reservoirs has increased rapidly and storage capacity
has decreased significantly. Total deposit rate of reservoirs in Taiwan is 28.6% averagely. The deposit
phenomena of reservoirs in southern Taiwan are most serious, total deposit rate even reaches 36.8%
(WRA, 2011).
2.2 Key issues of water resources adaptation
Based on the analysis of historical data, it shows that variation of annual rainfall in Taiwan has
dramatically changed. For instance, annual rainfall in 2005 is about 3,568 mm, whereas annual rainfall
in 2002 is 1,572 mm only. In addition, the variations of annual precipitation keep increasing, and the
period between wet-year and dry-year becomes shortened. Under the threats of climate change impacts,
the uncertainties of precipitation results in that the circumstances of natural water resources is difficult
to predict. Therefore, “development and conservation of water resources” is in advanced priority of
adaptation for water resources.
Watersheds in Taiwan have been highly developed. Due to increasing intensity of rainfall, the
soil erosion in catchments accelerate, abundant sediment strikes the capacity of reservoirs. As a result,
“effective water supply” is also in advanced priority of adaptation for water resources.
The demand of water, which includes agricultural consumption, domestic consumption and
industrial consumption, is very likely to increase due to impacts of climate change. Limited water
supply cannot satisfy the increasing demand; in consequence, “water demand” is the third advanced
issue of adaptation for water resources.
Besides domestic water resources, national water footprint can be gained indirectly by
import of agricultural and industrial products from foreign countries. Hence, “import and export of
water resources” is also taken into account for water resources adaptation. However, the water stress is
the main problem in near future; the first three key issues are addressed here. Figure 1 illustrates the
four key issues of adaptation for water resources in Taiwan.
Water Resources Adaptation

Evaptranspiration Precipitation
Issue Development and

I conservation of Surface water Groundwater
water resources
Sea Water storage in Assessing water

reservoirs in rivers
Issue Assessing water
II Effective in reservoirs
Groundwater
water supply abstraction
Water
purification
Water
transportation
Issue
Issue IV Import and
III Water Industrial Agricultural Domestic
export
demand consumption consumption consumption
Foreign
countries
Figure 1 Key issues of adaptation for water resources in Taiwan
3. Climate change scenarios
Global climate change is high dependence on emission of greenhouse gases (GHGs). In

different emission scenarios (A1, A1B, A2, B1 and B2) developed by IPCC, projections of future
temperatures shows significant differences. Considering characteristic of water resources in Taiwan
and the uncertainties of future climate, the adaptation actions are executed by stages and be reviewed
and revised periodically. At the first stage, target period is set as 2020~2039, and baseline period is set
as 1980~1999. The climate change projections are estimated, using the IPCC-developed scenario A1B,
which is considered by the international science community as the most likely future outcome.
Through on the studies from historical records, there are two possible hydrological changes
including: (1) increase of variations of annual rainfall, (2) increase of rainfall in wet season and
decrease of rainfall in dry season. In 24 General Circulation Models (GCMs), 9 of them perform the
later hydrological change. As a result, they are chosen to be the scenarios of hydrological change in
Taiwan for the following studies.
Precipitation changes, comparing with the one of baseline in the important river basins,
including Tamsui river, Tsengwen river and Kaoping river, are projected (figure 2). The increasing
rates of precipitation are 4%~17%, 4%~20% and 12%~82% in Northern area, Middle area and
Southern area of Taiwan, respectively.
Tamsui river basin Kaoping river basin

Rainfall (mm)
Rainfall (mm)
Baseline A1B Baseline
A1B
Tsengwen river basin

Rainfall (mm)
Baseline A1B
Figure 2 Baseline and projected rainfall in Taiwan important river basins
4. Risk assessment of water resources
4.1 Climate change impacts on water resources

The four climate change factors, which are (1) raising temperature, (2) decreasing total rain
days, (3) increasing rainfall intensity and (4) rising sea level, impact on water resources through series
studies. The impacts on three key issues of water resources and their 9 sub-issues with respect to four
climate change factors are listed in table 2.
By assessing the four climate change factors with 11 sub-issues, 9 main impacts on water
resources are identified (figure 3). “Raising temperature” mainly leads increasing demand of water
resources and reducing amount of surface water storage. “Decreasing total precipitation days per year”
mainly decreases the flexibility of water resource operation. “Increasing rainfall intensity” impacts on
water turbidity leading to decreasing retainable water and influences water accessibility. “Raising sea
level” mainly results in the decreasing amount of fresh groundwater. Figure 3 illustrates the linkage of
16 main impacts and 9 sub-issues of water resources. Those issues are exmained using water resources
system analysis to identified the vulnerbilities for four water resources operation systems in Taiwan.
Table 2 Analysis of climate change impacts on water resources
4 climate change factors

Increasing
3 key issues and 11 sub-issues Rising Decreasing Rising
rainfall
temperature total rain days sea level
intensity
amount of amount of - -
Development
Surface water surface water surface water
and
storage storage
conservation of
- amount of - amount of
water
Groundwater groundwater groundwater
resources
storage storage
amount of amount of water turbidity -
Water storage
water storage water storage and
in reservoirs
in reservoirs in reservoirs sedimentation
Water - - water turbidity -
accessibility
Effective water of reservoirs
supply Water - amount of water turbidity -
accessibility stream flow
of rivers
Water water quality water quality water turbidity -
purification
amount of amount of - -
Agricultural
agricultural irrigation water
consumption
water demand demand
amount of - - -
Domestic
Water demand domestic water
consumption
demand
amount of - - -
Industrial
industrial water
consumption
demand
Climate change factors

Increasing Decreasing
Rising sea level Rising temperature
rainfall intensity total rain days
Increasing Increasing Decreasing Increasing

Intrusion of saltwater
sedimentation turbidity stream flow evapotranspiration
4 7 8 Water quality
deterioration
13
5
10 11
1 2 3 6 9 12 14 15 16
1 2 3 Water 4 Water 5 Assessing 6 7 8 9

Surface Ground- Water Agricultural Domestic Industrial
accessibility accessibility water in
water water of reservoirs purification consumption consumption consumption
of rivers rivers
16 main impacts of climate change

9 main sub-issues of water resources
Figure 3 The linkage of 16 main impacts and 9 main sub-issues of water sources
4.2 Assessment of hazard, vulnerability, and risk
Figure 4 shows the conceptual rules of risk assessment for water resources under climate change
scenarios. The definition and approaches are presented as following.
1. Risk
(1) Definition: Risk is combination of hazard of climate factors and vulnerability of water
resources. Risk= hazard x vulnerability
(2) Evaluated rule: Based on matrix of hazard and vulnerability to evaluate the risk.
2. Hazard
(1) Definition: Hazard is the possibility of significant impacts caused by climate change.
(2) Evaluated rule:
A. High hazard: Before the target period, the possibility of climate change impacts
causing significant hazards is high.
B. Medium hazard: Before the target period, climate change impacts cause hazards, but
its significance is need follow-up studies.
C. Low hazard: Before the target period, the possibility of climate change impacts
causing significant hazards is low.
3. Vulnerability
(1) Definition: Vulnerability indicates integrated status of sensitivity, exposure and resilience of
water resources under climate change impacts.
(2) Evaluated rule:
A. High vulnerability: Under climate change impacts, the influenced extent and scale
increase significantly, and no adaptive measures are implemented yet. The resilience ability needs to
be strengthened.
B. Medium vulnerability: Under climate change impacts, the influenced extent and scale
increase significantly, but adaptive measures have been implemented. The resilience ability has be
strengthened.
C. Low vulnerability: Under climate change impacts, the influenced extent and scale do
not increase significantly.
Hazard
Climate change
Risk= hazard x vulnerability factors
High possibility
Low possibility
of significant
of significant
Possible hazards
 Hazard: the possibility of

need follow-up
hazards
hazards
studies
significant impacts
 Vulnerability: the consequences
of impacts
H M L
Vulnerability Water resources
adaptation
Influenced extent and scale
increase, no adaptive
measures are implemented
H H H M
increase, adaptive measures
have been implemented
M H M L
do not increase
L M L L
Risk
Figure 4 The conceptual rules of risk assessment for water resources under climate change scenarios
After evaluation, there are 3 sub-issues facing high risks under climate change. The first issue is
the decreasing of surface water resource. Under the most possible hydrological scenarios, the
precipitation become more concentrated, and the change rate of annual rainfall increase significantly.
Thus, it is highly possible that available surface water resources decrease. Surface water resources in
Taiwan have 93% contribution of total usable water resources. If surface water resources decrease
significantly, the usable water resources in Taiwan will decrease as well. As a result, new or integrated
water resources need to be developed in order to alleviate the impacts caused by climate change.
Second is the decreasing of reservoir storage. Decreasing total rainfall days indicates that
rainfall become more concentrated, and the rainfall amount of stroms increase significantly. In
addition, the increasing of rainfall intensity leads increasing sedimentation of reservoirs. In
consequence, it is highly possible that the storage of reservoirs will decrease. Reservoirs are the main
infrastructure to balance water storage in wet and dry seasons. If the water-sorage fuction of reservoirs
reduced significantly, capability and flexibility of water supply in Taiwan will be reduced as well.
Parts of reservoirs have reinforce dredging work and conjunctive utilization. The water-sorage fuction
of reservoirs need to be strengthened constantly.
Third is the increasing of agricultural water demand. Rising temperature indicates more
evapotranspiration, and the irrigation water demand for crops is bound to increase. As a result, it is
very possible that agricultural water demand will increase. Agricultural consumption accounts for
more than 70% of total. If agricultural water demand cannot be satisfied, it will influence the yield of
crops.
5. Climate change action plans
5.1 National adaptation framework

“National Climate Change Adaptation Policy Framework” has set 4 adaptation strategies (figure
5) for 4 key issues (sect. 2.2). WRA plans to promote “National Adaptation Action Plans for Water
Resources” coordinating related units, including Environmental Protection Administration (EPA),
Council of Agriculture (COA), Ministry of Economic Affairs (MOEA), Ministry of the Interior (MOI)
and National Science Council (NSC), to draw up corresponding plans based on the adaptation
strategies for water resources.
Responsible
National Climate Change Adaptation Policy authorities
CEPD
Responsible
Action Plans for Water Resources Adaptation
authorities
WRA
Related authorities
Development and conservation of water resources
Strategy
1 Taking sustainability as the highest guiding principle for the nation's
water resources management, and also focusing on protecting the aqua WRA, EPA, COA, MOEA
environment.
Effective water supply

Strategy
Reviewing water resources management from the supply side and WRA, EPA, COA, MOEA, NSC
2 reinforcing the efficient usage of water resources.
Water demand
Strategy
Establishing a database of the regional total quantity of water supply
3 system and reviewing water resource total quantity control policies from WRA, MOI, COA, MOEA
the demand side in order to reinforce the efficiency of water resources.
Import and export of water sources

Strategy
Promoting water sustainability in line with the United Nations' water
4 footprint concept. WRA, MOEA, NSC
Figure 5 Naitonal adaptation strategies of water resources and related authorities
5.2 Action plans for water resources
4 adaptation strategies and 14 corresponding measures for water resources are set in “National
Climate Change Adaptation Policy Framework” (sect. 1; table 1), and related authorities based on the
measures set 66 adaptation action plans. Based on assessment results of water resources risks under
climate change impact (sect. 4), 9 action plans for the 3 sub-issues facing high risks and 4 important
integrated action plans have advanced priorities to execute. Table 3 lists the 13 action plans and the
responsible authorities.
Table 3 13 action plans having advanced priorities for water resources adaptation
Responsible
Classification NO. Action plans
authorities
Enhancement of mechanism and strategies of water resources
Decreasing 1 WRA
under climate chaege
surface water
Improvement and management of water environment and master
resource 2 EPA
plan
Reinforcement and improvement projects for water-storage
3 WRA
Decreasing constructions (phase II))
reservoir 4 Remediation project for reservior and catchment of Shimen WRA
storage Demostration project for low impact development planning in
5 WRA
catchment of reservior
6 Improvement of management of water right WRA
Planning the promotion policy for efficiency of uses in
Increasing of 7 COA
argricultural water resources
water demand Saving agricultural consumption by promoting dryland irrigation
for agriculture 8 COA
piping and modernizing water-saving facilities
Promoting water-saving irrigation and adjusting farming
9 COA
mechanism in subsidence area of Changhua and Yunlin counties
Strategy planning and promotion of water fee for sustainable
10 WRA
operation of drinking water industry
Compilation of the storm water storage standard and legislation of
Important 11 MOI
retention facilities of buildings
integrated plans
12 Planning of regulation pools during rezoning agricultural land COA
Establishment of legislation and reward mechanismof the water
13 WRA
re-use
6. Conclusion
Climate change is a worldwide and ongoing event that all countries around the world confront
and suffer different levels of damage. In the face of climate change, the proper courses of adaptive
actions should be to taken in order to reinforce adaptive capacities before a natural disaster strikes, and
to recover from a disaster as soon as possible (CEPD, 2013). CEPD has drawn up the National
Climate Change Adaptation Policy Framework which serves as the main basis for each department's
adaptation work.
Taiwan has unique rainfall patterns and fragile geological characteristics. As a result, draw up
adaptive policies responding to climate change is very important, especially in water resources sectors.
WRA and related government departments have assessed the degree of hazard, vulnerability, and risk
of climate change impacts. In addition, the adaptation action plans are established for the high-risk
areas. Through the following researches, the adaptation action plans would be built up and adjusted
periodically to enhance the adaptation capacity of water resources in Taiwan.
References
Council for Economic Planning and Development , 2012, Adaptation Strategy to Climate Change in
Taiwan.
IPCC, 2007, Fourth Assessment Report, Climate Change (AR4).
Water Resource Agency, MOEA., 2013, Assessment Report of Water Resources under Climate
Change (Draft).
Water Resource Agency, MOEA., 2011, Reservoir Sediment Releasing Countermeasures. Cope with
Climate Change (2/2) (in Chinese).
Water Resource Agency, MOEA., 2013, The Second Stage Management Project of Climate Change
Impacts and Adaptation on Water Environment (3/4) (in Chinese).
Water Resource Agency, MOEA., 2011, Water Resources Agency, Ministrey of Economic
Affairs, Profile.
Session 2
THE FUTURE
B-04
Water saving effects of the simplified surge flow and alternate dry
furrow methods in Uzbekistan
Junya Onishi*, Paluashova Ghavharay**, Hiroshi Ikeura
* Rural Development Division, Japan International Research Center for Agricultural Sciences
** Research Institute of Irrigation and Water Problem
ABSTRACT
In Central Asia, large-scale irrigation development was conducted in the Amudarya and Syrdarya river
basins, which were previously steppe or desert areas, starting in the 1960s during the Soviet Union era.
Irrigation development enabled the farming of cotton and wheat in such areas but caused secondary
salinization in the irrigated land. Secondary salinization occurs mainly because of salt input to
farmlands from irrigation water and rising groundwater levels caused by excessive irrigation. In the
Republic of Uzbekistan, furrow irrigation is widely practiced, but there is a tendency for excessive
irrigation, resulting in high infiltration loss. Therefore, the area of salinized farmlands in Uzbekistan is
the largest in Central Asia and measures against salinization are urgently needed.
In this study, water-saving irrigation methods based on furrow irrigation were introduced to reduce
excessive irrigation. A field experiment comparing the water-saving effects of the following 5 methods
was conducted: (1) furrow irrigation at a farmer’s discretion, (2) furrow irrigation with an appropriate
amount of water, (3) alternate dry furrow method (ADF), (4) simplified surge flow method (Surge), and
(5) a combination of ADF and Surge (ADF + Surge).
The water infiltration showed higher reductions in Method (3) than in Method (1), but no such clear
reductions were observed in the other methods. The Surge method was not found to be effective because
of the low infiltration rate of the field.
Keywords Secondary salinization, Furrow irrigation, Excessive irrigation, Water-saving irrigation
1. Introduction
In Central Asia in the 1960s, during the Soviet Union era, large-scale irrigation development
was conducted in the Amudarya and Syrdarya river basins, which had previously been steppe or desert
areas. Especially, the Republic of Uzbekistan (hereinafter Uzbekistan), which had an arid or semiarid
climate, was assigned the role of cotton production on the developed farmland. This enabled the
farming of cotton and wheat on dry land.
The large-scale development of irrigated agriculture relied on furrow irrigation. An advantage of
furrow irrigation is that farmers with minimal capital investment for facilities and materials can
perform it. Therefore, even after independence from the Soviet Union, furrow irrigation is still widely
practiced. However, the drawback of furrow irrigation is the large infiltration loss and its promotion of
secondary salinization caused by rising groundwater levels in Central Asia.
Uzbekistan
Google Map
Figure 1. Location of Uzbekistan
Secondary salinization is a critical issue for sustainable agriculture in arid and semiarid land.
The main causes are salt input to the field with irrigation water and rising groundwater levels due to
excessive irrigation that promote capillary rising. In the Central Asian plains, most soils are naturally
salted and potentially dangerous in their secondary salinization development (Yu. I. Shirokova et al.
2006). Table 1 shows the area of irrigated farmland that is affected by salinization in Central Asia.
Uzbekistan has the largest area of irrigated land in the region, and half of them were salinized caused
by secondary salinization.
Table 1 Salinized land in Central Asia

Irrigated area Area affected by salinization
Country
(ha) (ha) (%)
Uzbekistan 4,280,600 2,140,500 50.1
Kyrgyzstan 1,077,100 124,300 11.5
Tajikistan 719,200 115,000 16.0
Kazakhstan 2,313,000 >763,290 >33.0
Turkmenistan 1,744,100 1,672,592 95.9
Central Asia 10,134,000 4,815,732 47.5
Source: World Bank (2003)
In order to mitigate salinization, efficient irrigation methods should be introduced. The proposed
irrigation methods are drip, sprinkler, and other high-efficiency irrigation methods. However, those
methods need high-pressure water delivery systems. Therefore, larger capital expenditures are
necessary when converting from furrow irrigation. In addition, new facilities require electricity and
other extra energy costs. However, considering the real economic conditions of farmers in Uzbekistan,
it is difficult to introduce these technologies because of a lack of investment funds. From the
viewpoint of feasibility, the improvement of present furrow irrigation is acceptable and realistic for
farmers.
In this study, in order to reduce the infiltration loss of furrow irrigation, a low-cost improvement
method based on furrow irrigation was introduced to actual farm fields in Uzbekistan and the
reduction in infiltration loss was examined.
2. Project site
The project site was selected considering the percentage of salinized land. The Syrdarya Region
is located approximately 120-km southwest of the city of Tashkent (Figure 2). Ninety-eight percent of
irrigated land is salinized in the region. The Syrdarya Region has a steppe climate. According to the
weather data from 2004 to 2012 obtained from two weather stations (Syrdarya and Yangiyer), the
maximum monthly average temperature is around 30°C and the minimum is around 0°C. The annual
precipitation is around 320 mm, and it is concentrated in the winter (Figure 3). Therefore, irrigation is
mainly necessary in the summer season.
The experimental site was selected in a field belonging to the Yangiobad Water Consumers
Association (hereinafter WCA) in Mirzabad District. Yangiobad WCA area has 2300 ha of salinized
land, and it is one of the largest salinized areas in the Syrdarya Region. Irrigation development was
carried out in 1920–1960, and furrow irrigation has been performed for a long time. The groundwater
level is relatively high, around -3.0 to -0.5 m.
Rain (mm) Temp (℃)
30.0 60.0
Temperature (℃) 25.0 50.0
Rainfall (mm)
20.0 40.0
Tashkent 15.0 30.0
Syrdarya region 10.0 20.0
5.0 10.0
0.0 0.0
Google Map 1 2 3 4 5 6 7 8 9 10 11 12 (Mon)
Figure 2. Location of Syrdarya region Figure 3. Average monthly temperature and

rainfall in the Syrdarya region (2004–2012)
3. Materials and Methods

3.1 Introduced irrigation methods
In order to clarify the effect of water saving by improved furrow irrigation methods, a field
experiment was carried out. The introduced irrigation methods were as follows.
(1) Alternate dry furrow method
The alternate dry furrow (hereinafter ADF) method involves irrigating alternate furrows (Figure
4), (FAO 1988). Water is supplied to two ridges from one furrow. The advantage of the ADF method
is that it reduces the amount of applied water and decreases infiltration loss by non-irrigated furrows
and lateral flow. Alan R. Mitchell et al. (1993) reported that 50% of irrigation water is saved by the
ADF method compared to the normal furrow method.
(2) Simplified surge flow

The surge flow method involves irrigating water intermittently instead of continuously. The
advantage of this method is that it decreases infiltration loss by reducing soil permeability with the
cyclic application of water. The water flow of the second water supply is faster than the first water
supply because of the reduction of permeability by the first water supply. The reduction in infiltration
is caused by four physical processes: consolidation due to soil particle migration and reorientation, air
entrapment, redistribution of water, and channel smoothing (Alan R. Mitchell et al. 1994).
The normal surge flow method is usually introduced using a butterfly valve and switch
controller to supply water in a series of pulses. However, considering the financial conditions of
farmers in Uzbekistan, the normal method is not acceptable because the costs of butterfly valves and
switch controllers are high. Therefore, in this study the surge flow method was simplified. The
simplified surge flow (hereinafter Surge) method involves irrigating twice with no equipment. The
first supply (1st Surge) extends up to 50 m along the furrow from the inlet by normal furrow irrigation,
while the second supply extends up to 100 m along the furrow from the inlet one day after the 1st water
supply from normal furrow irrigation (2nd Surge) (Figure 5). This method is more acceptable for
farmers, because no additional investment or complicated management is required.
Normal method Normal method

0m 50 m 100 m
Normal irrigation
Root zone
Large infiltration loss
Surge flow method

ADF method
0m 50 m 100 m
1st water supply
2nd water supply
Root zone
st nd
1 infiltration loss 2 infiltration loss
Figure 4. ADF method Figure 5. Simplified surge method
(3) Combined ADF and Surge

It is expected that the combination of the two irrigation methods mentioned above (ADF +
Surge) will have an even greater effect.
3.2 Design of the experimental site

The experiment was conducted in actual farm field from the 23rd to 25th of July 2012. Figure 6
shows a scheme of the experimental fields. Five treatments were tested to compare the effect of saving
water; 1) every furrow irrigation at the farmer’s discretion (Control), 2) every furrow irrigation with a
proper amount of irrigation water (Conventional), 3) the ADF method (ADF), 4) the Surge Flow
method in every furrow (Surge), 5) combined ADF and Surge (ADF + Surge). Each experimental plot
consisted of eight parabolic furrows and ridges (The standard width of a furrow was 0.4 m and that of
ridges was 0.5 m). The width of one test plot was 7.2 m (furrow 0.4 m + ridge 0.5 m = 0.9 m, 0.9 m ×
8 = 7.2 m) and the length was 100 m. The furrow gradient was 0.1% (1/1,000). There were three
replicates. Before starting the experiment, ordinary faming treatments, such as plowing, harrow
sowing, etc., were conducted in all test plots.
Inlet Earth canal
Weir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Combination (ADF + Surge)

Conventional
Conventional
Conventional
Control
Control
Control
Surge
Surge
Surge
ADF
ADF
ADF
100m
Concrete canal
Next plot Border 1.8 m Plot 7.2 m
0.9 m
2 Furrows 8 Furrows
No irrigation Irrigation according to treatment
Figure 6. Design of the experimental plot
3.3 Amount of irrigation water

The amount of irrigation water in the Control was decided by the farmer, while in the other plots
it was calculated based on soil moisture-retention characteristics and cotton root system analysis
conducted in 2011 in the Yangiobad WCA area. In this experiment, it was assumed that the yield was
not reduced by moisture stress; therefore, the Readily Available Moisture (RAM pF 1.8–3.0) was
adopted for calculation. The analysis of soil moisture retention was conducted using a soil sample
from the experimental site. The soil had a measured pF of 1.6 to 3.2 based on 3 measurements using
the pressurized membrane method. The results of the analysis were that the field capacity and
depletion of moisture content for normal growth were 0.443 cm3/cm3 and 0.267 cm3/cm3, respectively.
The average RAM in 3 measurements was 0.176 cm3/cm3. The root survey data from 105 days after
sowing were used for this experiment, 0–20 cm was considered the critical soil layer, and the ratio of
moisture consumption of the critical soil layer was 92.2%. The appropriate amount of irrigation water
was decided using the total readily available moisture (TRAM) and leaching requirement (LR) (FAO
1994). TRAM and LR are calculated using formulas (1) and (2), respectively.
1
TRAM = ( f C − M L ) × D × (1),
Cp
where TRAM is in millimeters; fc is the field capacity (pF 1.8); ML is the depletion of moisture content
for normal growth (pF 3.0); D is the thickness of critical soil layer; and Cp is the percentage of
moisture consumption of the critical soil layer.
ECw
LR = (2)
5( ECe) − ECw
where LR is the minimum leaching requirement needed to control salts within the tolerance (ECe) of
the crop with an ordinary surface method of irrigation; ECw is the electrical conductivity of the
applied irrigation water in dS/m (value in Syrdarya experimental fields: 1.36 dS/m); and ECe is the
electrical conductivity of soil saturation extract in dS/m (maximum ECe to obtain 100% of the cotton
yield: 7.7 dS/m).
The calculated results of (1) and (2) are applied in formula (3).
TRAM
AW = (3),
1 − LR
where AW is the depth of applied water (mm); TRAM is the total readily available moisture (mm); and
LR is the minimum leaching requirement needed to control salts within the tolerance (ECe) of the crop
with the ordinary surface method of irrigation.
According to the above calculation, the appropriate amount of irrigation water is 39.6 mm
during the experiment after 88–90 days of sowing. In order to supply the necessary amount of water to
the ridges, the water depth in the furrow was calculated by formula (4).
Wi
ID f = AW × (4),
Wf
where IDf is the irrigation water depth in the furrow (mm); AW is the depth of the applied water (mm);
Wi is the width of the ridge and furrow supplied water by one furrow (cm); and Wf is the width of the
furrow (cm).
The collection factor (Wi/Wf) for the conventional and Surge treatments was 90/40, and for the
ADF and ADF + Surge treatments, it was 140/40. Table 2 shows the necessary depth of irrigation
water at the end of the furrow in each treatment.
Table 2. Necessary depth of irrigation water

No Plot Applied water (mm)
(1) Control Farmers decision
(2) Conventional 89
(3) ADF 139
(4) Surge 89
(5) Combination 139
3.3 Measurement and analysis

Irrigation water discharge in each plot was measured using a rectangular weir. Water discharge
from three furrows in each plot was measured using a triangular weir.
Furrow infiltration was measured before irrigation by using a Marriott tank in each plot. In the
case of ADF and ADF + Surge, measurement was conducted without flooding in the left and right side
of the furrow. In the case of Surge Flow and ADF + Surge, measurement was conducted before and
after the 1st water supply. The furrow infiltration test was continued for 240 min.
From the results of the measurements, an infiltration curve was derived using the Kostiakov
equation (Formula (5)).
D f = ct in (5),
where Df is the infiltration water depth (mm); c and n are constants of the Kostiakov equation; and
t i is the elapsed time for infiltration (min).
The advance and recession times of irrigation water was measured at every 10 m from the inlet
to 100 m in three furrows in each plot. The water advance equation and water recession equation are
shown in formulas (6) and (7). From the results of measuring the water advance and recession, the
constant penetration of water advance and recession was determined using formulas (6) and (7).
t a = α 1 Lα β 1 (6),
where t a is the elapsed time of water advance (s); α 1 and β 1 are constants of the water advance
equation; and La is the distance water advanced from the inlet in the furrow (m).
t r = α 2 Lr + β 2 (7),
where t r is the elapsed time of water recession (s); α 2 and β 2 are constants of the water recession
equation; and Lr is the position of water recession measured in the furrow.
The infiltration water depth of one furrow and ridge at every 1 m in the furrow was calculated
by formula (8), which uses the results of formula (5), (6), and (7).
t − t 
n
D f = c r a  (8),
 60 
where Df is the infiltration water depth (mm).
The amount of infiltration water at every 1 m was calculated using formula (9). The total
amount of infiltration water is the sum of Sj from 1 to 100 m.
 D fj + D fj +1  
  × W 
 2  
Sj =  (9),
1000
where Sj is the amount of infiltration water at an arbitrary 1 m of furrow (m3); Dfj is the infiltration
water depth at an arbitrary point in the furrow (mm); Dfj+1 is the infiltration water depth at an arbitrary
point in the furrow + 1 m (mm); and W is the infiltration width (mm).
4. Results and Discussion
4.1 Amount of irrigation water and time taken for irrigation
The average amount of water applied to three furrows in each plot is shown in Figure 7, and the
average irrigation times of three furrows in each plot are shown in Figure 8. The amount of water and
time for Surge and ADF + Surge are the total of the 1st and 2nd water supplies. The water amount and
irrigation time of the control was the largest in all trials. These results imply that the farmer used a lot
of water and spent time in irrigation. Excess water included (1) water that flowed away from the open
end of the furrow and (2) increased infiltration water caused by the longer irrigation time.
(m 3 ) Trial 1 Trial 2 Trial 3

10
8
6
4
2
0
Cont Conv ADF Surge A+S Cont Conv ADF Surge A+S Cont Conv ADF Surge A+S
No.1 No.2 No.3 No.4 No.5 No.10 No.6 No.7 No.8 No.9 No.13 No.14 No.15 No.11 No.12
Figure 7. Average amount of irrigation water in one furrow
(min) Trial 1 Trial 2 Trial 3

80
60
40
20
0
Figure 8. Average irrigation time in one furrow
4.2 Furrow infiltration

The relationship between the cumulative infiltration and elapsed time is shown in Figure 9. The
cumulative infiltration after 240 min was around 30 mm, but it was under 20 mm in some plots such as
No.5 (ADF + Surge) and No.12 (ADF + Surge), while No.2 (Conventional) had cumulative infiltration
of around 40 mm. The amount of infiltrated water was less and the permeability of the experimental
field seemed to be low. In this field, a compacted layer with a bulk density of 1.65–1.70 g/cm3 was
observed 20-cm below the ground surface, which decreased permeability. According to this result,
furrow infiltration is uneven in the field. Further, this is presumed to occur depending on the location
in the field, such as being close to the drainage or the center.
In the Surge method, a reduction in furrow infiltration was expected after the 1st water supply,
but in the case of No.5 (ADF + Surge) an increase in furrow infiltration after the 1st water supply
occurred, which was not expected. The other plots showed a reducing effect after the 1st water supply.
On the basis of the above results, furrow infiltration seemed to be highly dependent on the
position in the field because the permeability of the field was not even.
Trial 1 Trial 2 Trial 3
(mm) No.1 Control No.10 Control No.13 Control
40
30 y = 2.4354x0.4221 y = 1.9005x0.509 y = 1.8339x0.5016
R² = 0.9444 R² = 0.9596 R² = 0.9556
20
10
0
0 40 80 120 160 200 240 0 40 80 120 160 200 240 0 40 80 120 160 200 240
No.2 Conventional No.6 Conventional No.14 Conventional
40 y = 1.6376x0.5864 y = 2.0495x0.5095
R² = 0.9975 R² = 0.9745
30 y = 1.7829x0.4581
R² = 0.9669
20
10
0
0 40 80 120 160 200 240 0 40 80 120 160 200 240 0 40 80 120 160 200 240
No.3 ADF No.7 ADF No.15 ADF
40 y = 1.9012x0.5242
R² = 0.9622
30 y = 1.8019x0.4687 y = 1.7499x0.4101
R² = 0.9942 R² = 0.9658
20
10
0
0 40 80 120 160 200 240 0 40 80 120 160 200 240 0 40 80 120 160 200 240
No.4 Surge No.8 Surge No.11 Surge
40 y= 3.368x0.4403 y = 2.3905x0.4805
R² = 0.9793 R² = 0.9661
30 y = 1.7049x0.4955
R² = 0.9627
20
y = 0.8977x0.6138
10 y = 1.3263x0.5035 R² = 0.9448 y = 1.5692x0.4572
R² = 0.9765 R² = 0.9721
0
0 40 80 120 160 200 240 0 40 80 120 160 200 240 0 40 80 120 160 200 240
No.5 ADF+Surge No.9 ADF+Surge No.12 ADF+Surge

40 y= 1.0158x0.5347 y = 0.7655x0.5111
y= 1.1647x0.451
R² = 0.9662 R² = 0.9851 R² = 0.9872
30 y = 1.7698x0.4735
y = 1.1255x0.4323 y = 1.3643x0.4634
R² = 0.9646 R² = 0.9592 R² = 0.9619
20
10
0
0 40 80 120 160 200 240 0 40 80 120 160 200 240 0 40 80 120 160 200 240
(min)
Before irrigation After irrigation
Figure 9. Cumulative infiltration and elapsed time
(s) No.1 Control No.10 Control No.13 Control
15000 y = 22.327x + 6950.2
R² = 0.2572
10000 y = 72.618x + 10089
y = 58.798x + 7560.5
R² = 0.6945 R² = 0.6792
5000 y = 12.7219 x1.0515 y = 7.0639 x1.1543 y = 11.7296 x1.0782
R² = 0.9876 R² = 0.9954 R² = 0.9741
0
0 50 100 0 50 100 0 50 100
No.2 Conventional No.6 Conventional No.14 Conventional
15000
y = 59.844x + 7447.6 y = 48.307x + 4539.2 y = 42.796x + 3644.4
R² = 0.7504 R² = 0.9199 R² = 0.9065
10000
5000
y = 10.871x1.1417 y = 14.744x1.0473 y = 4.595x1.2675
R² = 0.9268 R² = 0.9975 R² = 0.985
0
0 50 100 0 50 100 0 50 100
No.3 ADF No.7 ADF No.15 ADF
15000
y = 42.895x + 6133.4 y = 48.199x + 3838.1 y = 24.588x + 5267.5
R² = 0.7856 R² = 0.7605 R² = 0.7116
10000
5000 y = 9.0723x1.1547 y = 7.4989x1.1249 y = 8.3263x1.2375

R² = 0.968 R² = 0.9902 R² = 0.9806
0
0 50 100 0 50 100 0 50 100
No.4 Surge 50m No.8 Surge 50m No.11 Surge 50m
15000
y = 14.908x + 2340.9 y = 17.254x + 4025.1
R² = 0.7373 R² = 0.0521
10000 y = 16.274x + 4375.5 y = 13.437x1.0824 y = 14.298x1.004
R² = 0.043 R² = 0.9896 R² = 0.9907
5000
y = 16.909x1.0394
R² = 0.9748
0
0 50 100 0 50 100 0 50 100
No.4 Surge 100m No.8 Surge 100m No.11 Surge 100m
15000
y = 22.765x + 4821.6 y = 40.294x + 5421.8 y = 45.252x + 6770.4
R² = 0.2421 R² = 0.8789 R² = 0.8495
10000
5000 y = 6.5853x1.082 y = 4.4387x1.2195

y = 2.6946x1.3689
R² = 0.9857 R² = 0.9823 R² = 0.9909
0
0 50 100 0 50 100 0 50 100
No.5 ADF+Surge 50m No.9 ADF+Surge 50m No.12 ADF+Surge 50m
15000 y = 13.261x + 2050 y = 31.362x + 2906
y = -13.314x + 3755.5
R² = 0.1939 R² = 0.8894
10000 R² = 0.0602
y = 14.866x0.892 y = 14.122x0.991
y = 13.52x0.931
R² = 0.9772 R² = 0.9994
5000 R² = 0.9948
0
0 50 100 0 50 100 0 50 100
No.5 ADF+Surge 100m No.9 ADF+Surge 100m No.12 ADF+Surge 100m

15000
y = 10.727x + 7169 y = 83.65x + 2838.3 y = 5.3064x + 5496.6
R² = 0.0434 R² = 0.8316 R² = 0.1054
10000
5000 y = 6.0291x1.1259 y = 4.2163x1.1144

y = 6.1802x1.0829
R² = 0.9686 R² = 0.99
R² = 0.9877
0
0 50 100 0 50 100 0 50 100
(m)
Water advance Water Recession
Figure 10. Water advance and recession and elapsed time
4.3 Water advance and recession
The flow velocity of the supplied water was not even. In order to compare the water advance
rate and water recession rate, furrows with a similar flow velocity were selected for the analysis. The
average water velocity obtained for 39 of the 45 furrows was 0.10 m3/min.
Figure 10 shows the relationship between the water advance, recession, and elapsed time. The
time for supplied water to reach 100 m is around 1000 to 2000 s. The time for the applied water to
disappear from the surface of the field was around 10,000 to 15,000 s in each plot. A 0.1% slope was
present from the inlet to the end of each furrow. This means that the inlet side was at a higher
elevation, and thus, the supplied water accumulated at the end of the furrow, and the water recession
time at the end of the furrow side was very high in almost all plots.
4.4 Infiltration loss
Figure 11 shows the calculated amount of water necessary for infiltration and infiltration loss in
each treatment. The necessary amount of water was considered as that when the root zone received
sufficient water along with the entire length of each furrow.

6.00
Infiltration water (m 3 )
4.00 0.07 0.29 0.05 0.31 2.26 0.03 0.05 0.02 0.22 0.74 0.18 0.12 0.01 0.07 0.03
2.00
3.57 3.57 3.57 3.57 3.57 3.57 3.57 3.57 3.57 3.57 3.57 3.57 3.57 3.57 3.57
0.00
Necessary infiltration (39.6 mm) Infiltration loss
Figure 11. Amount of water for necessary infiltration and subsequent infiltration loss
Infiltration loss in all trials was low except for ADF + Surge (No.5, 9). One factor behind this
result may be the low permeability of the field. For the ADF method, reduced infiltration loss was
obtained in all trials; thus, alternate irrigation appears effective for water saving. In Surge and ADF +
Surge, the reduction effect was obtained only in Trial 3. In No.5, furrow infiltration after the 1st water
supply increased, indicating no reduced permeability. As a result, irrigating twice led to large
infiltration losses. Other plots (Nos. 4, 8, 9, 11, and 12) showed reduced permeability after the 1st
water supply, but this was not found in Trial 3. It is presumed that the reducing effect of the Surge
method is related not only to infiltration but also to water advance and recession.
5. Conclusion
In this experiment, the ADF method reduced infiltration loss in all trials. However, the Surge
and ADF + Surge methods did not reduce infiltration loss. It is presumed that the Surge method was
strongly affected by the field permeability. If the Surge method is applied in a field with low
permeability, irrigating twice might simply increase infiltration loss.
According to this result, it is important to implement enough surveys to understand field soil
physical properties before introducing the Surge method. If the field permeability is not low, reduced
infiltration loss by the Surge method can be expected. The ADF + Surge might require additional labor
such as closing or opening furrows for ADF and irrigating twice for Surge, and thus, it is important
that future research focus on improving the methods to make them more practical for use by farmers.
Acknowledgments
This study was conducted with a subsidy from the Ministry of Agriculture, Forestry and
Fisheries in Japan. We also express our appreciation for all people who supported this study.
References
Alan R. Mitchell and Karen Stevenson (1994), “Surge flow and alternating furrow irrigation of
peppermint to conserve water.” Central Oregon Agricultural Research Center Annual Report
1993, AES OSU, Special Report 930, 79-87.
Alan R. Mitchell, Joy E. Light and Tera Page (1993), “Alternate and alternating furrow irrigation of
peppermint to minimize nitrate leaching.” Central Oregon Agriculture Research Center Annual
Report 1990-1991, AES OSU, Special Report 922 29-36.
FAO (1994), Irrigation and drainage paper No. 29, Water quality for agriculture,
http://www.fao.org/docrep/003/t0234e/T0234E03.htm#ch2.4.2, Accessed 13 September.
FAO (1988), Training manual No. 5, Irrigation water management: irrigation methods,
http://www.fao.org/docrep/s8684e/s8684e04.htm#chapter%203.%20furrow%20irrigation,
Accessed 13 September.
World Bank, Julia Bucknall, Irina Klytchnikova, Julian Lampietti, Mark Lundell, Monica Scatasta and
Mike Thurman (2003), Irrigation in Central Asia: Social, economic and environmental
considerations, 1-12.
Yu.I. Shirokova and A.N. Morozov (2006), “Salinity of irrigated land of Uzbekistan: causes and
present stage.” Springer, Sabkha Ecosystems Volume II: West and Central Asia, 249-259.
THE FUTURE
B-07
Expansion of Leased Paddy Land and Crisis of Sustainability of

Water User Associations in Japan
Hajime Tanji*, Katsuhiro Sakurai**, Ataru Nakamura***, Hirohide Kiri*, Tetsuo

Nakaya*
*National Institute for Rural Engineering, NARO, Japan
**Faculti of Economics, Rissho University
***Graduate School of Frontier Sciences, The University of Tokyo
ABSTRACT
In 2000, the MAFF drastically expanded its farm land ownership policy to include leased farmland by
revising the Agricultural Land Law. The change of agricultural policy strongly influenced farmland
ownership and the size of land cultivated by each farmer. Irrigation schemes are managed by Water
User Associations (WUAs) based on the Land Improvement Law. The main problem is that policies
under laws promoted by two laws are rarely harmonized．The authors studied the situation of the Aichi
WUA, and selected the Ikeda Water Management Unit (WMU) with approximately 200 land owners.
The total area cultivated by the largest 6 farmers (the top 2% of the cultivators) occupies more than
60% of Ikeda’s WMU. WUA is established by the Land Improvement Law (1949) during a farmland
reform was carried out between 1947 and 1950. Under this farmland reform, all farmers cultivate
about their own farmland. The Land Improvement Law assumes that most cultivators are also land
owners. At present, this condition is not satisfied, because most cultivators are tenant farmers. This
changed condition makes it difficult to maintain the decision making systems of a WUA. In the case of
the Ikeda WMU, one farmer manages his water every three days because they must manage the water
of more than 10 areas containing outside of Ikeda WMU. This situation makes it difficult to harmonize
water management with neighboring paddy fields. Irrigation water management is facing a
sustainability crisis. To solve this problem, the Land Improvement Law must inevitably be amended.
Keywords: Gini coefficient, inverse Lorentz curve, Land Improvement Law, water user association,
farmland concentration
1. Background of study
In Japan, the population peaked in 2004. The working age population, people from 15 to 64, is
decreasing rapidly. Many industries have responded to this by changing their labor policy. Some
companies moved factories out of Japan. In the case of agriculture, the MAFF changed its farmland
policy in 2000 to concentrate farmland in the hands of core farmers by expanding rental land. The
2005 and 2010 World census of Agriculture and Forestry in Japan showed an expansion of rental farm
land. This expansion process has been discussed by many agricultural economists. But there has been
little discussion of the influence of this policy on irrigation water management by WUA. In this
paper, the authors discuss this influence through a case study of the Aichi Water User Association
(AWUA).
2. Study area
The main study area was selected as the Aichi Irrigation Scheme shown in Figure 1. The Aichi
IS was constructed from 1957 to 1961 as the first phase, then as a second phase, reconstructed from
2005 to 2007 to repair and improve its facilities. The irrigated area in 2012 was 13,584 ha. This
scheme includes the Maiko Dam and a 100km long main canal, secondary canals, tertiary canals, farm
ponds, and quaternary canals. The Aichi IS was originally designed only for irrigation. The growth of
the economy created demand for water for industrial and domestic use. The main canal of the Aichi IS
is used for irrigation, industrial and domestic water supply.
Among its typical facilities, the main canal, secondary canals and tertiary canals are owned by
the WRA (Water Resources Authority, a government organization). The Maiko Dam, Kaneyama
Headworks and the main canal are managed by the WRA. The Kaneyama Headworks is a main intake
weir located at the upstream end of the main canal. Secondary and tertiary canals are managed by the
Aichi Water Users Association. The Aichi WUA was set up in 1952 to construct and manage the Aichi
Irrigation Scheme. Among management facilities of the Aichi WUA, secondary canals are managed
directly by the Aichi WUA. Management of tertiary canals is entrusted to the WMU (Water
Management Unit) of farmers by the Aichi WUA. Farm ponds and the quaternary canals are managed
by farmers groups. Owners of these facilities are farmers groups formed at each farm pond. Before the
Aichi Irrigation Scheme, each farmers group owned their farm pond and managed the pond and the
quaternary canals with fees they collected. The Aichi IS was designed to transmit water to these farm
ponds. After construction of the Aichi Irrigation Scheme, farmers groups continued to perform water
management. They stopped collecting fees and began paying fees to the Aichi IS because water
management by farmers groups became very easy and simple.
A detailed survey of the Ikeda branch area was done. This area is managed by the Ikeda WMU.
Makio Dam
Kiso River
Inuyama HW Kanayama HW
Main canal
Nagoya city
Ikeda branch area

HW(Headworks)
Ponds
Supplied area
Irrigated land
5km
Figure 1 Outline of the Aichi Irrigation Scheme
3. History of expansion of concentration of farmland
After establishment of the Agricultural Land Law in 1954, the MAFF considered the small size
of cultivated areas to be the main cause of low productivity of agriculture and an obstacle to
mechanization. Many procedures and projects have been adopted to expand the size of farmland. But
cultivated areas have not expanded. Instead of expanding the size of their farms, farmers have
obtained additional income by taking side jobs. Opportunities to take such side jobs were greatly
increased by new factories opening in rural areas. Expansion of the size of farms has progressed, but
only very slowly. Stockbreeding farmers first expanded the number of livestock. Paddy fields grew
more slowly because the invention of new mechanized methods has helped drastically decrease
farming labor. Many paddy farmers are called Sunday farmers, because they can grow rice by working
only on Sundays. This stable condition of rice farming has also stabilized the management of the
WUAs.
This trend changed in the 1990s when the population almost stopped growing. Since the
Uruguay Round in 1994, the farm gate price of rice has decreased. Many factories moved abroad from
Japan. A new policy of some kind should be implemented.
Since the adoption of the farmland expansion policy in 2000 by the MAFF, two sets of census data,
2005 World census of Agriculture and Forestry in Japan and another in 2010 have been announced.
After the 2010 data were announced, many researches discussed the expansion of concentration using
the 2000 data as the benchmark.
Figure 2 shows the classification of farmers under the census. Farmers are classified by their
cultivated areas and the total selling price of their agricultural production. A farmer with cultivated
area larger than 30a and total selling price larger than 500,000 yen is classified as a “selling farmer”. A
farmer smaller than a selling farmer but with cultivated area larger than 10a and total selling price
larger than 150,000 yen is classified into a “self-sufficient farmer”. A farmer with cultivated area less
than 5a is omitted from the census.
There is another category, “a management body”, which refers to a group of farmers who manage and cultivate
agricultural land as a group.
Cultivated area
Selling
farmer
30a
Self-sufficient
farmer
10a
No farmer
landowner
5a
Out of survey
150,000 500,000 yen

Total sold price
Figure 2 Classification of farmers in a census
From 2005 to 2010, the area cultivated by selling farmers and management bodies increased
from 1.9 to 2.2ha. Among them, 43.3% cultivated more than 5ha in 2005 and 51.4% cultivated more
than 5ha in 2010.
In the Tokai District, where the Aichi Irrigation Scheme is located, the number of selling farmers in
2010 decreased to 53% of the number in 1990. The number of non-cultivating land owners in 2010
increased to 168% of that in 1990. Self-sufficient farmers leased about 40% of their owned land and
non-cultivating landowners leased about 90% of their owned land.
Now the actual land ownership situation is a bit complicated. Figure 3 shows the percentages of
contracted work by selling farmers and management bodies in Tokai district. Almost a quarter of the
work is contracted. Figure 3 shows that the size of farmland is not related to cultivators, but to
farmland managers.
% Whole Japan Gifu Aichi Mie

45
40
35
30
25
20
15
10
5
0
Narsring Paddling Tranplanting Pest control Clutivation Drying
Figure 3 Percentage of entrusted in each work (2010 census)
3. History of water user associations
The origin of present irrigation projects is historically based on two laws. One is the Arable
Land Adjustment Law established in 1899. This law deals with farmland readjustment to expand the
size of plots and to improve irrigation and drainage conditions. To describe this law more precisely,
the original law of 1899 did not include any provisions related to irrigation and drainage. The revised
law of 1909 first stipulated land readjustment for improving irrigation and drainage. The other is the
Irrigation Association Law established in 1908. This law deals with the adjustment of water use for
irrigation. For wider areas, the River Law provided for water rights and the adjustment of irrigation
associations. The Ministry of the Interior legislated river training based on the River Law. Irrigation
projects were supported under the Irrigation Association Law enacted by the Ministry of the Interior.
The Irrigation Association Law is one umbrella law under the River Law. On the other hand, the
Ministry of Agriculture supported projects under the Arable Land Adjustment Law. Until the end of
the Second World War, there were many large landowners and landless farmers. The objects of arable
land adjustment projects were landowners. The 1899 law defined the objects of projects as landowner
groups and the condition for a state project was the agreement of more than two thirds of the
landowners, the total area and the total value of the area respectively. In 1909, this condition changed.
Requirements concerning the number of landowners decreased to only more than half.
After, the Second World War, Japan was temporarily governed by GHQ. GHQ aimed to prevent
Japan from ever starting another war. GHQ intended to divide up core organizations and groups that
had existed before the war. Large land owners were one of its targets. In 1946, the Agricultural Land
Reform Law was established under the leadership of GHQ. This law required landowners to sell all
their farmland except 1 ha at a predetermined low price. Land reform was done 16 times from 1947 to
1950. Through this process, the percentage of landless farmers changed from between 40 and 48% to
9%. But, the Agricultural Land Law was established in 1952, just after the land reform.
Another target of GHQ was the Ministry of the Interior. Before the war, the political power of
the Ministry of the Interior was much stronger than that of other ministries. GHQ considered that
power should be balanced among ministries. In 1949, the Land Improvement Law was established.
This law replaced the Irrigation Association Law and Areal Land Adjustment Law. The object of land
adjustment and irrigation was only one organization; the water user association. The combination of
the two laws has merits and demerits.
3.1 Merits
Irrigation and drainage projects can only start by request of a WUA. A WUA became the only
type of organization related to farmland improvement including middle or large scale irrigation and
drainage. WUAs manage irrigation and drainage projects and land readjustment based on the Land
Improvement Law. Normally, the biggest project promoted by a WUA is an irrigation and drainage
project concerning main facilities, dams, headworks and main canals. A WUA established under the
Land Improvement Law should be called a Land Improvement District (LID). The Aichi WUA is
definitely called the Land Improvement District for the Aichi Irrigation Scheme. But this combined
system is rarely adopted in other countries because of the high cost of land readjustment. If irrigation
and drainage projects and land readjustment projects can be promoted by the same organization,
synchronization of construction of three factors, irrigation, drainage and the large shape of farmland
will effectively improve the productivity of farmland. But this condition is not always governed by
WUAs because irrigation and drainage projects are hierarchical and land readjustment projects are not
hierarchical. In the irrigated area of the Aichi Irrigation Scheme, tertiary canals not constructed by the
AIS have been governed by secondary canals constructed by the AIS. Land readjustment projects
should be synchronized with irrigation and drainage projects at the tertiary canal level. But there is no
need for synchronization with an irrigation and drainage project at the main and secondary canal level.
The Aichi WUA has promoted small land readjustment projects linked to irrigation and drainage
projects at the tertiary canal level. This did not cover the whole irrigated area in the Aichi WUA.
Similar small projects had been promoted by the local government. In this paper, the authors do not
use the name, LID, but rather use WUA because of the commonality of English words about irrigation
and drainage.
3.2 Demerits
Almost no paper has discussed the demerits of this combined law system. In the authors’
opinion, this combined system has caused a confusion concerning public goods. Normal definition of a
public good is non-excludable and non-rivalrous. In the case of a public good, the market system does
not work well. Drainage projects have non-exclusive characteristics. This fact shows that a drainage
project does not supply a public good but rather a common good (common-pool resources). But
irrigation projects supply water under exclusive and rivalrous conditions. In some other cases, market
failures are known to have occurred. It is difficult to for the market to supply a decreasing cost
industry because of its high initial construction cost. Construction of a large irrigation scheme requires
a huge initial cost. This fact shows that an irrigation project is one type of decreasing cost industry.
According to this idea, the size of a governmental irrigation project has a 3,000ha lower limit of its
irrigated area, and a prefectural irrigation project has a 500ha lower limit. Governmental and
prefectural projects are only applied when market failure is expected. This normal interpretation of a
decreasing cost industry for irrigation projects is apparent at a glance. But considering the procedures
under the combined laws, the situation seems a bit different.
Under the Land Improvement Law, a WUA has the power to limit the property of its members.
This power is given so it can perform land readjustment. In an arable land readjustment project, the
agreement of all members of the intended area is actually very difficult to obtain. To overcome this
difficulty, even if only a few people in the intended area do not agree on land readjustment, the WUA
has the power under the law to start and complete the project. To confirm this purpose, the rate of
agreement is very important. Under the Arable Land Readjustment Law, 2/3 or 1/2 of the number of
land owners, the area and the value of land is adopted as that necessary for this confirmation as shown
in Table 1.
But curiously, this strength is not required as a result of conditions in a decreasing cost industry.
An irrigation project is characteristically excludable and rivalrous. Irrigation water can be treated as a
private good. Nevertheless, a 2/3 agreement is a condition required to start an irrigation project. It is
very difficult to treat irrigation water as a public good. The authors think that a 2/3 agreement
Table 1 Laws related to farmland adjustment

Land Total Total
Year Law Cultivator Entity
Owner Area Value
1899 Arable land readjustment >2/3 >2/3 >2/3 Individuals, Group
1909 Arable land readjustment >1/2 >2/3 >2/3 Individuals, Group, Organization
1949 Land Improvement >2/3 Organization (WUA)
Table 2 a and b Laws related to farmland adjustment and irrigation and drainage
Before the war
Project Irrigation and drainage project Areal land readjustment project
Law Irrigation association law Areal land readjustment law
Agreement none >2/3 of area, value, landowners
Compulsory power none tax collection
Organization Irrigation association land readjustment association
After the war

Project Irrigation and drainage project Areal land readjustment project
Law land improvement law
Agreement >2/3 of cultivators
Compulsory power limit property
Organization Water user association
condition is not essential for an irrigation project. If the initial cost condition of a decreasing cost
industry can be cleared, even a private company can supply irrigation water. If you understand the
relation of JR Tokai with the Japan National Railroad Authority concerning Shinkansen trains, you
can agree with this opinion.
Table 2 shows differences between the law and the organization system before and after the war.
4. Change of frameworks between land owners and cultivators
For farmland reform, basic framework of laws is the relation between land ownership and
cultivation which had prevailed for the past 100 years in Japan. Figure 4 shows change of farmland
system. Before the war, a few land owners owned huge areas of farmland. These farmlands were
cultivated by landless farmers. After the war, as a result of the agricultural land reform, land owned by
large scale land owners was divided and all cultivators became land owners. The size variance of land
owned by farmers is small and the size of their farms is also small. The present WUA was established
by farmers of that period. The Land Improvement Law is based on a tacit understanding that all
farmers own farms of similar size and that all cultivators own their farmland. Agricultural land reform
gave them strong political power but it resulted in the problem of the small size of land. Recently,
selling farmers and management bodies use larger areas of farmland. But as shown in Figure 3, land
owners or cultivators do not own or cultivate such large land areas. The authors propose a new concept
of land manager to describe this relationship. In Figure 4, concentration of farmland is expressed as
the activities of land managers.
Land owner
Before the war

Cultivator Cultivator Cultivator
Land owner Land owner Land owner
After the war
Land owner Land owner Land owner
Recently
Land manager
Figure 4 Framework of change between land owners and cultivators
5. Case study of Ikeda WMU
On March 2012, the authors interviewed farmers and cultivators in the Ikeda WMU. Farmers
were selected based on their cultivated area. Six farmers were interviewed. Among these six, one
farmer has no cultivated land in the Ikeda WMU area. This chapter describes the result for the other 5
farmers. The Ikeda WMU includes about 200 land owners. The total area of the Ikeda WMU is 2500a
(25ha). Only 100 of the 200 land owners now cultivate land. The other 100 landowners lease their
farmland. Because of this leased land, there is a very large variance in the cultivated areas of the 100
cultivators. Based on this fact, interviews focused on only large-scale cultivators. The biggest 5
cultivators were selected for the interviews. Figure 5 shows the areas cultivated by and ages of the 5
farmers. The 5 farmers cultivated a total of 1660a. This means that the remaining 840a is cultivated by
the remaining 100 farmers. This distribution shows a high concentration of farmland. Figure 5 shows
that, all the land cultivated by the 5 farmers is located in the Ikeda WMU, and that the farmer with
cultivated area of 90a (70 years) cultivates 10a outside of the Ikeda WMU. The biggest cultivator
farms 500a (5ha). According to the census survey method, a farmer cultivating an area of 5ha is not
treated as concentration case. This fact indicates that the results of census underestimate the
concentration of farmland. Its actual influence on the WUA is more radical. Figure 6 shows the above
mentioned past and future expected farmland concentration in the Ikeda WMU.
500 65
840 74
54
320 55
70
90 Others
400
350
Figure5 Breakdown of cultivated land in the Ikeda WMU (unit a)

(Numbers in the legend are ages of farmers.)
After the war 210 farmers Land owner Cultivator
100 farmers Land owner Cultivator
Present Land manager Cultivator

6 farmers Land owner
No
cultivator
10 farmers Land owner Cultivator
Future
3 farmers Land owner Land manager Cultivator
No
cultivator
Figure 6 Change in structure of farmers in Ikeda WMU
6. Basic problem with past analysis of farmland concentration
Accumulation of farmland has been studied by many researchers. Uchida (2012) analyzed the
results of the 2010 World census of Agriculture and Forestry in Japan. His analysis showed that 21%
of farmland bigger than 10 ha and 15% of that bigger than 20 ha in the Tokai district where the Aichi
Irrigation Scheme is located. The study of farmland accumulation faces the following problems.
1) The 2010 census data is not provided in its raw format. In the case of data for farmland size,
data is provided as the frequency of a histogram just as shown in the figure. Frequency histogram
data are data averaged by each width of a histogram. This data treatment loses information
concerning the variance of the data. From histogram data, the authors tried to restore the original
distribution. But all attempts to recover the original data are biased to underestimate the variance of
the original data.
2) Many research projects focused on accumulated farmland bigger than 10 ha or 20 ha. But for
WUA, these criteria are nonsense.
3) For WUA, the 2/3 condition is very important. But past analysis did not account for this
problem.
Inequity is normally expressed as a Lorenz curve and Gini coefficient. In a Lorenz curve, all
data are arranged in ascending order. And a Gini coefficient shows the distance from equity condition.
To discuss the agreement of a project, data should be arranged in descending order. Therefore, an
inverse Lorenz curve should be used to discuss the inequity of cultivators.
Figure 7 shows the inverse Lorenz curve of selling farmers. The curved line 0Z shows this distribution.
The X axis is the cumulative percentage of number of farmers and the Y axis is the cumulative
percentage of farmland area. No variance can be expressed as a straight line 0Z. The tacit
understanding behind the Land Improvement Law can be expressed as the straight line 0Z under the
condition that all farmers are selling farmers. In that case, the 2/3 of cultivators condition can be
expressed as XC1-XC2. And the 2/3 farmland area condition can be expressed as the line Y-Y.
Though this condition is not described in the Land Improvement Law, this condition is tacitly
understood to be fulfilled. If the inverse Lorenz curve is 0Z and the tacit understanding is not satisfied,
the 2/3 farmland area condition is equal to 30% of cultivators by XB1-XB2 line.
From here, the X axis is assumed to express all farmers in the WMU. If 50% of farmers stopped
cultivating, no variance line could be expressed as the straight line 0ZZ and the inverse Lorenz curve
could be expressed as a curve 0ZZ. According to the XA1-XA2 line, the 2/3 area condition is equal to
15% of land owners.
This analysis shows the following points.
1) The influence of accumulation of land on the WUA is very large because the tacit understanding
is not satisfied. For the 2/3 condition, by number of landowners, by number of cultivators, and
number of land managers and land area should be considered.
2) “A failure rate of tacit understanding” can be expressed as the ratio of XA1-XA2 to Y-Y in
Figure 7.
Cumulative area
XA2 XB2 ZZ XC2 Z

100
CumTArea2Per
80
Y Y
40
20
20
60
60
40 0
XA１ 20XB1
0 40
XC1
60 80
CumPerPercent
100
Cumulative person
Figure 7 Inverse Lorentz curves of cultivators and farmland
7. Recommendation of a solution
The present situation shown by interviews in the Ikeda WMU and by census data indicates the
failure of the tacit understanding. The 2/3 condition under the Land Improvement Law should be
revised to avoid this failure. A key point is that the 2/3 condition should be applied not to the number
of people but to the size of the area.
For a large irrigation and drainage project, the 2/3 condition is not essential. It is possible to
separate land adjustment from large scale irrigation and drainage projects. Considering foreign
irrigation and drainage schemes, separated systems are more common.
At Tagajo city in Miyagi prefecture, the land improvement district vanished. At present, the
local government plays the role of WUA. This fact shows that a new organization for land
readjustment and irrigation and drainage projects will be required as the number of farmers decreases.
The recommendable new organization or new system is not clear now. In any case, revision of the 2/3
condition will be a major issue.
8. Conclusion
Since 2000, the accumulation of farmland has expanded. This has resulted in the failure of the
tacit understanding concerning conditions for agreement in a WUA. The failure has been destroying
the fiscal foundations of the WUA and increasing the difficulty of renewal projects. According to the
results of a case study in the Aichi WUA, accumulation has a stronger influence on WUAs than on
farmers. Considering the expansion speed, little time remains to recover from this failure.
The authors proposed “A failure rate of tacit understanding”, to evaluate the failure condition of
the WUA. The risk to each WUA can be evaluated by this index. But to find a solution to prevent
failure, there is a limit to the efforts which can be made by each WUA. A general solution based on
the revision of the Land Improvement Law will be unavoidable.
Acknowledgement
The authors thank the Aichi Water User Association for its generous cooperation with the
interviews of water managers and farmers. This research was supported by “Demand Oriented
Irrigation Service Study (Research leader: Dr. Toshiaki Iida)” of Service Science, Solutions and
Foundation Integrated Research Program of RESTEX of the Japan Science and Technology Agency.
References
Uchida Takio, 2012, “Trends and issue of accumulation of management farmland”, Norinkinyuu,
Vol.11, 720-731 (in Japanese)
Session 3
THE FUTURE
C-01
Water Management at Large-Sized, Sub-Irrigation-Installed

Paddy Fields
NAKAMURA Kazumasa, KOHIYAMA Masayuki and UNOKI Keiji
In extensive rice-producing regions in Hokkaido, the total area of rice paddy fields is intended to
increase to at least 30 ha per farming household, in that farm lands are expected to be consolidated to
meet the decrease in the number of succeeding farmers. In rice farming areas where farmers manage
such a large-scale farming, the size of individual field parcels is being extended to 2 ha and
sub-irrigation facilities are being installed to make work on site more efficient. In this study, we
surveyed the water levels of undergrounds and submerged fields at those improved large-sized paddy
fields. We also summarized the characteristics of irrigation water demand including irrigation
requirements and water intake strengths at each stage of rice growth. Additionally, we conducted
water distribution simulations that reflect the summarized characteristics of irrigation water demand.
These simulations have clarified points to bear in mind in managing water distribution when
collecting water intensively within a single distribution system, for example, at the initial growth stage
of rice that is directly sowed on submerged fields.
1. Introduction
The Food, Agriculture and Rural Areas Basic Plan formulated by the Japanese government in
2010 states that improvement of paddy fields is necessary in order to facilitate rotational cropping
toward helping to increase Japan's food self-sufficiency. According to this plan, paddy fields in Japan
have been improved by the addition of underground pipes that are used not only for water discharge
but also for water supply and groundwater level control, with an eye to introducing direct sowing
culture of rice as well as to labor-saving production of wheat, soybeans and vegetables. The farm
management scale in the rice-producing regions of Hokkaido is greater than in anywhere else in Japan,
and the area of rice paddy field per farming household is expected to increase to at least 30 hectares in
some of Hokkaido's rice-producing regions in the near future, due to the ongoing acceleration of
demographic aging. In these regions, the above-mentioned underground pipes have been developed
and the size of individual paddy field lots is being expanded for the purpose of increasing farm work
efficiency.
In the regions undergoing such development and expansion, there may be some changes in the
use of irrigation water, as follows: a) The maximum flow rate of water intake will increase because of
the larger field lots and the introduction of subirrigation. b) After the introduction of subirrigation,
irrigation requirements will change in those paddy fields where direct sowing culture is conducted,
because soil and water are managed in a different way than in the fields where transplanting culture is
implemented. c) In those paddy fields where improvements include the construction of water
distribution pipelines as part of irrigation canals, the operation of a faucet at a parcel of paddy field
will affect the rate of water discharge from other faucets in neighboring parcels, for hydraulic reasons.
These three factors that cause changes in the irrigation water use pattern can adversely affect the
impartiality of water distribution to the parcels of paddy fields in an area covered by a single water
distribution system.
Since paddy field improvement is expected to be implemented in many regions in the future, it
is necessary to anticipate possible problems in the management of water distribution and to prepare
solutions to these problems. For the purpose of addressing these problems and proposing solutions, the
researchers in this study conducted a field survey to look into local water demand characteristics and
performed water distribution simulations on the basis of the water demand assumed for the future.
2. Research on water management in paddy fields
2.1 Methodology
2.1.1 Paddy fields used for research
Research on water management was conducted during a period of paddy-field rice farming from
May through September in 2011 at three parcels of paddy fields that are close to each other in the
Moseushi town, Hokkaido (figure 1). These three parcels of paddy fields are hereinafter called Parcels
A, C and D. The processes of cultivation in each parcel and the dates each process was applied are
shown in table 1. Improvement of these parcels was completed in 2009, when government projects
for land readjustment and the construction of a subirrigation system were implemented (figure 2). In
these projects, the irrigation canal and the drainage canal adjacent to each parcel were replaced by
conduits buried under farm roads for the purpose of saving the labor necessary for weeding and other
maintenance work
.
Parcel C Parcel D Parcel A
1.5ha 1.5ha 1.5ha
0 100km
Moseushi town
0 100m
Figure 1 Location of the parcels for study
Table 1 Cultivation processes and the dates each process was applied (2011)
Parcels Parcel A Parcel C Parcel D
Transplanting Direct sowing in Direct sowing in
Cultivation technique
culture submerged fields submerged fields
Soil puddling May. 10-24 No operation No operation
Initial water intake No operation May 29 May 29
Transplanting/Sowing May 29 May 30 May 30
Drying up of paddy fields Jun. 26 Jun. 29 Jun. 29
Re-submerging of paddy Jul. 4-5 Jul. 4-5 Jul. 4-5
fields
Water management for Jul. 4 - Aug. 12 Aug. 10-24 Aug. 10-24
protein control
Harvesting Sep. 10-24 Oct. 12 Oct. 12
After the completion of the improvement work, unique cultivation techniques were applied to
these parcels of paddy fields. Specifically, Parcels C and D have been used for the pioneering adoption
of direct sowing culture in submerged paddy fields without soil puddling, with the aim of realizing
labor-saving cultivation of rice. In Parcels C and D, the groundwater level has been raised or lowered
after the emergence of rice ears to remove excess nitrogen from the soil and, thus, to control the
nitrogen absorption by rice plants in order to enhance the flavor of the rice. In this study, the artificial
adjustment of the groundwater level mentioned above is called "water management for protein
control."
Concrete mass
Faucet
Drainage conduit Feeder road (5m wide)
Water distribution Conduit Water supply

conduit to soil
Bulkhead
(Height-
adjustable) Underdrainage
pipe
Legend
Water pipe
Irrigation control unit Irrigation control unit
Sluice for groundwater
level adjustment
Surface drain outlet

Parcel surface
Water distribution conduit
Movable slit for
Drainage conduit controlling the
Underdrainage conduit groundwater level
in the parcel
Groundwater
level in the parcel
Water flow
Plan view of the parcels
Sluice for groundwater

level adjustment
Figure 2 Irrigation and drainage system for parcels
Decisions were made by the farmers concerned regarding whether they would use surface
irrigation or underground irrigation. The farmers also determined the rate of water intake by adjusting
the opening of faucets.
2.1.2 Observation
The groundwater level and the depth of ponding were respectively observed at four points in
each of the three parcels. The observation points are shown in figure 3. At each of the observation
points in each parcel, measurements were taken at intervals of 10 minutes. Regarding both the
Underdrainage pipes
Table 2 Periods defined for the purpose of calculating the

1/4 quantity of irrigation water
Period Description
I Transplanting culture: From the time for water intake for
1/4 soil puddling through transplanting
Direct sowing: The time for initial water intake
II From the time immediately after transplanting or sowing
through the time for draining and drying up paddy fields
1/4
III From the time for re-submerging paddy fields through the
time of rice-ear emergence
IV From the time immediately after rice-ear emergence
1/4
through the end of the irrigation period (Aug. 31)
Observation points for ponding depth

Observation points for groundwater level
Figure 3 Observation points in a parcel
groundwater level and the depth of ponding, an average value of the four measurement points was
calculated, and each average value was also expressed in terms of an elevation. The groundwater
gauge used for measurement was placed 1 m below the paddy field surface.
A tipping bucket rain gauge was set near the parcels for measurement of the amount of rainfall
every 10 minutes.
2.1.3 Calculation of the quantity of water intake

The quantity of water intake at each parcel was calculated for each of four periods that are
characterized by different farm operations and growth stages of the rice plant. The four periods are
explained in table 2. Because of the structural characteristics of the irrigation control unit shown in
figure 2, it is not possible to install a flowmeter on the control unit in order to directly measure the
flow rate of water intake. Thus, to obtain the quantity of water intake, the amount of change in the
groundwater level was multiplied by the soil porosity and the product was added to the amount of
change in the depth of ponding. The soil porosity used in the calculation was based on the data
collected on the days with rainfall in and after September. To calculate soil porosity, the amount of
rainfall (mm) was divided by the increase in the groundwater level (mm).
2.2 Results
2.2.1 The depth of ponding and the groundwater level
Figure 4 shows data on the daily rainfall as well as on the groundwater level and the depth of
ponding at each parcel of paddy field for study. This figure contains information on the water
management implemented by farmers and the irrigation method (surface irrigation, underground
irrigation or the concurrent use of those two methods).
Decisions are made separately in different parcels as to the irrigation method that is optimum at
different times during the cultivation period. In Parcel A, both surface irrigation and underground
irrigation were used during the time of soil puddling, and surface irrigation alone was mainly used at
other times. On the other hand, underground irrigation was mainly used in Parcels C and D, where rice
seeds were directly sown in the submerged paddy fields. If surface irrigation is used for supplying
water to the fields initially and also after seed sowing, horizontal water flow in the fields will cause the
rice seeds to migrate.
Precipitation
(mm/d)
Ground water table (cm) Ground water table (cm) Ground water table (cm)
Both Surface Both Under- Surface
Ponding depth (cm)

Parcel A
ground
Sluice left
open
Sluice left
open Water
Re-submerging
Water intake for after drying up management for
soil puddling paddy fields protein control
Underground Underground Both Under- Parcel C
Ponding depth (cm)

ground
Sluice left Sluice left

open open
Shallow water Water Water

management intake management for
Re-submerging after Water intake protein control
Initial water intake drying up paddy fields
Ponding depth (cm)

Underground Underground Both Parcel D
Sluice left
Sluice left Average
open
open groundwater level
Shallow water Water Average depth of
management intake ponding
Re-submerging after Water Height of paddy
Initial water intake drying up paddy fields intake field surface
Figure 4 Temporal changes in the groundwater level and the depth of ponding, and
irrigation methods (Underground: Underground irrigation; Surface: Surface irrigation; Both:
Underground and surface irrigation)
Figure 4 shows the changes in the groundwater level specific to the water management that is
different from the typical water management in other paddy fields in the same local area. Specifically,
shallow water management after sowing in Parcels C and D, and water management for protein
control in Parcels A and C resulted in the groundwater level changes shown in figure 4. In the latter
half of the shallow water management period, the lowest groundwater level was 60 cm below the
surface. A large amount of irrigation water should be taken when it is necessary to raise the
groundwater level from that depth to the surface of the paddy fields in one water intake operation.
Because shallow water management is conducted for the purpose of dampening the soil surface on a
periodic basis, when the lowest groundwater level is kept at a level much closer to the ground surface,
the amount of water intake in one water intake operation can be reduced. Regarding the behavior of
the groundwater level in the paddy fields where water management for protein control was
implemented, the rise in the groundwater level was relatively small at the fourth water intake in Parcel
A and in the fifth water intake in Parcel C. To determine the most appropriate number of days for
water discharge and irrigation during the period of water management for protein control, both the
behavior of the groundwater level shown in figure 4 and the effectiveness in controlling nitrogen
absorption should be considered.
Figure 5 shows the changes in the depth of ponding and the groundwater level during each
irrigation period in Parcels A and C when water was discharged via culverts by opening sluices. In
Parcel A, where soil puddling was conducted, lowering of the groundwater level did not greatly affect
the depth of ponding. In Parcel C, where soil puddling was not conducted, lowering of the
groundwater level caused a decrease in the depth of ponding and eventually resulted in the absence of
Parcel A （With soil puddling）
Ground water table（cm） -10 10
Ponding depth（cm）
0 0
Ground w ater table
10 Ponding depth -10
20 Height of paddyfield surface -20
30 -30
40 -40
50 -50
7/11 0:00 7/11 12:00 7/12 0:00 7/12 12:00 7/13 0:00
Parcel C （Without soil puddling）

Ground water table（cm）
-10 10
Ponding depth（cm）
0 0
10 -10
20 -20
Ground w ater table
30 -30
Ponding depth
40 Height of paddyfield surface -40
50 -50
6/29 0:00 6/29 12:00 6/30 0:00 6/30 12:00 7/1 0:00
Figure 5 Changes in the groundwater level and the depth of ponding when the sluice is open
ponding in the paddy field. This difference in the state of paddy fields in Parcel A and Parcel C was
caused by the difference in permeability of the plow pan. Although rice cultivation generally uses soil
puddling at present, cultivation methods that do not depend on soil puddling may become widely used
in the future. Then, the vertical permeability of the plow pan in paddy fields is likely to increase.
2.2.2 Characteristics of the demand for irrigation water at each parcel

The amount of water intake per period and the rate of water intake in one water intake operation
are shown in table 3 and figure 6 respectively. Each period is defined as shown in table 2. Due to the
unstable weather conditions in early spring in 2011, the timing of the first water intake was delayed in
Parcels C and D. The first water intake in these parcels was conducted shortly after the water intake
for soil puddling in Parcel A. Under normal weather conditions, the first water intake for direct sowing
culture in submerged paddy fields is usually conducted about one week before the water intake for soil
puddling in the paddy fields for transplanting culture. The amount of water for initial irrigation was
136.2mm in Parcel A, where transplanting culture was adopted. This amount is within the range of the
standard water quantity for soil puddling in Japan: 100 ~180mm (Shimura, et al., 1987). In Parcels C
and D where seeds were directly sown in submerged paddy fields, the amount of water intake was
smaller than in Parcel A for transplanting culture, but the water flow rate was greater in Parcels C and
D due to intensive water intake in a short time. In Parcels C and D, water intake was completed in a
short time because it was necessary for the farmers concerned to finish the water intake before sunset
so that they would be able to stop the supply of water immediately after visually confirming that the
surface of the paddy fields was entirely submerged. Thus, if intensive water intake is conducted at
many paddy fields in the same local area within a short period of time, the stability of the water supply
will be adversely affected.
In Period II, the amount of water intake was greater in the submerged paddy fields where seeds
were directly sown than in the paddy fields where transplanting culture was conducted. As in the case
with the initial water intake, a large quantity of water was also supplied in a short time to the paddy
fields when shallow water management was conducted, as shown in figure 6. Normally, the amount of
water intake becomes large from the beginning to the middle of May for soil puddling, and also
between mid-May and early June for initial water intake and shallow water management (Hokkaido
Agricultural Development and Promotion Association, 2005). In regions where both transplanting
culture and direct sowing culture are conducted, the water supply to paddy fields is expected to peak in
a period from the middle through the end of May, when soil puddling, initial water intake and shallow
Table 3 Amount of water intake by period
Parcel Period I Period II Period III Period IV
Parcel A 136.2mm 73.7mm 114.0mm 158.9mm
Parcel C 45.0mm 365.7mm 249.1mm 253.7mm
Parcel D 47.8mm 379.6mm 316.1mm －
water management are implemented at different paddy fields. In this regard, hydraulic studies are
necessary to see whether water distribution via pipelines is ensured in the peak period of water supply.
In Period III, water management for protein control was implemented in Parcels A and C.
Although the timing of the implementation was different between Parcel A and Parcel C, the quantity
of water intake in the implementation period was great in both parcels. When water management for
protein control becomes widely used in the future, the demand for irrigation water is likely to peak
during the period of water management for protein control. In light of this, it is necessary to collect
data on the effectiveness and the necessity of various measures before examining possible irrigation
planning methods.
Regarding each of the farm operations in both direct sowing culture and transplanting culture,
the changes in the groundwater level and in the depth of ponding were used for calculating the average
amount of water intake in one water intake operation, the duration of water intake and the intervals
between water intake operations as summarized in table 4.
3. Water distribution simulation
3.1 Outline of the simulation method

In water distribution simulations, input data include the time of day when water supply to paddy
fields is started, the amount of water intake needed for each parcel, the number of parcels where water
is supplied at the same time and the extent to which each faucet is opened. Simulation results are used
for estimating the temporal changes in the discharge rate of water from each faucet and the time of day
when water intake is completed at each of the parcels served by a single water distribution system.
These results are useful for comparing the flow rate of water from a faucet and the time necessary for
completing water intake among different parcels. When the simulation results indicate that the water
flow rate and/or the time when water intake is completed greatly differ among different parcels, and
that farmers are consequently forced to operate faucets at an inconvenient time of day such as at night,
a simulation can be conducted again by using different input data on the number of parcels to which
water is supplied on the same day or other items, and the simulation results are used for examining
measures for improving the management of water distribution.
A program developed by Hasegawa et al. (2001) was used for the water distribution simulations
in this study. The flow of a series of calculations is shown in figure 7. First, conditions are set
regarding the time for starting calculations and the percentages of faucet opening, and steady flow
analysis is conducted for calculating the quantity of water discharged from the faucet at each parcel.
Next, the calculation conditions are modified to calculate the cumulative quantity of water intake
every 30 minutes and to turn off the faucet at a parcel where the cumulative total has exceeded the
predetermined necessary amount of water intake. Then the quantity of water supplied to each parcel in
the following 30 minutes is calculated. The same calculation process is repeated until the faucets at all
parcels are turned off. The steady flow analysis in this program is performed by means of the
Takakuwa’s node energy potential method (Naito, 1988).
05/21 05/31 06/10 06/20 06/30 07/10 07/20 07/30 08/09 08/19 08/29
Precipitation（mm/d）
20
日降水量(mm)
40
60
80
100
Intake rate(mm/h)
15
代かき圃場 A
A Parcel
Water 取水強度(mm/h)
Puddling
10
移植
Transplanting Water management for protein control
低タンパク対策
0
05/21 05/31 06/10 06/20 06/30 07/10 07/20 07/30 08/09 08/19 08/29
Intake rate(mm/h)
15
圃場 C
C Parcel
Shallow water management
Water取水強度(mm/h)
浅水管理
10
Initial water Water management for
初期入水
intake protein
低タンパク対策control
5
0
05/21 05/31 06/10 06/20 06/30 07/10 07/20 07/30 08/09 08/19 08/29
Intake rate(mm/h)
15
浅水管理
Shallow water 圃場 D
D Parcel
Water 取水強度(mm/h)
management
10
Initial water
初期入水
intake
5
0
05/21 05/31 06/10 06/20 06/30 07/10 07/20 07/30 08/09 08/19 08/29
Figure 6 Rate and duration of water intake per water intake

Table 4 Quantity and intervals: Water intake for various farms
Cultivation Farm operation Days between water Duration of each Quantity of Required
method intake operations water intake water depth of
operation intake water
H L/s/ha mm
Transplanting Rough soil puddling － 18.2 9.0 58.8
culture
Transplanting Finishing soil puddling The day after rough soil 20.3 6.1 44.9
culture puddling
Transplanting Transplanting 3 days after finishing soil 21.2 3.9 29.6
culture puddling
Direct sowing Initial water intake － 10.5 12.3 46.4
culture
Direct sowing Shallow water management 6 14.9 22.5 93.0
culture
Both culture Middle stage irrigation 5 14.8 5.9 34.7
types
Both culture Re-submerging after drying up － 31.4 15.2 112.9
types paddy fields
Both culture Water management for protein 2 40.7 2.7 35.5
types control (average)
Both culture Water management for protein 2 35.7 5.4 73.8
types control (maximum)
3.2 Parcels served by a single water distribution system and the conditions assumed
for the simulations
The paddy field zone used for the simulations, which is served by a single water distribution
system, is in Moseushi town and has an area of 27.8 ha. This zone is composed of 11 parcels of paddy
fields. The elevation and other conditions of each parcel are shown in figure 8.
The period used for the simulations is between the middle and the end of May, when the
demand for irrigation water is expected to increase.
The input data were determined by giving consideration to the actual crop conversion rate and
farming plans in the paddy field zone. Because the current crop conversion rate according to the
farming plans is 25%, the simulations were conducted on the assumption that 8 parcels are used for
rice cultivation and 3 parcels are used for dry-field crop cultivation. The percentages of transplanting
culture and direct sowing culture in this zone were assumed on the basis of the future farming program
Start
Irrigation control unit
Input data on conduits and faucet opening

branch pipeline
Calculate the flow rate every 30 minutes

No.11 No.10 No.9 No.8 No.7
A 4.8ha A 2.4ha A 2.4ha A 2.4ha A 2.4ha
Calculate the quantity of water supplied to a parcel every 30
H 37.14m H 37.36m H 37.36m H 37.55m H 37.55m
minutes
branch pipeline
Calculate the cumulative quantity of water intake at each
parcel
branch pipeline
No.6 No.5 No.4 No.3 No.2 No.1
Yes Are calculations for 24 A 2.1ha A 2.1ha A 2.1ha A 2.3ha A 2.4ha A 2.4ha
hours finished ? H 37.05m H 37.05m H 37.36m H 37.36m H 37.60m H 37.60m
No
Turn off the faucet for each parcel where water intake is branch pipeline
completed
main canal (open channel)

End
Figure 8 Parcels of paddy fields used for

Figure 7 Flow of calculation in the simulation the simulations
and on interviews with farmers. It is expected that direct sowing culture will be adopted at about 50%
of the paddy fields even when this culture method is maximally utilized in this zone. In view of this, it
was assumed for the simulations that 50% of the paddy fields (i.e., 4 parcels) used transplanting
culture and the other 50% (4 parcels) used direct sowing culture.
An upper limit is put on the quantity of water diverted from a main irrigation canal to each
water distribution system. If water is supplied to all the eight parcels of paddy fields in one water
intake operation, the upper limit is exceeded. Thus, it is not reasonable to assume a situation in which
water is simultaneously distributed to the eight parcels. Instead, it was assumed that water intake is
conducted at four parcels in one water intake operation so that the quantity of water diverted to the
water distribution system would be slightly less than the upper limit.
Because the simulations targeted the time of year when the demand for irrigation water
increases substantially, it was necessary to understand in advance the correlations between various
types of farm operations conducted at each parcel and an increase in the water demand. In this study, a
combination of a particular farm operation that requires water intake and the number of parcels where
that farm operation is conducted is called a water intake pattern. The series of studies described above
indicate that the water demand increases during the initial irrigation period. For the purpose of the
simulations, it was assumed that the duration of the initial irrigation was 15 days. The number of days
when water is supplied to the eight parcels during the 15 days was simulated by taking into account
the number of parcels that adopt either transplanting culture or direct sowing culture. The simulation
results are shown in table 5 in terms of two water intake patterns. Middle Stage Irrigation (M.S.I.) in
Table 5 Assumed water intake patterns for the eight parcels
Water intake pattern 1
Day in the 15-day period
Cultivation method 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
F.I. S.W.M. S.W.M.
Direct sowing culture
F.I. S.W.M. S.W.M.
F.I. S.W.M. S.W.M.
F.I. S.W.M. S.W.M.
Transplanting culture R.P. F.P. T.P. M .S.I
.
Transplanting culture R.P. F.P. T.P. M .S.I.
Amount of water intake (L/s/ha) 49.2 36.0 24.4 0 0 15.6 90.0 0 0 0 23.6 0 90.0 0 0
Water intake pattern 2

Day in the 15-day period
Cultivation method 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
F.I. S.W.M. S.W.M.
F.I. S.W.M. S.W.M.
F.I. S.W.M. S.W.M.
F.I. S.W.M. S.W.M.
Amount of water intake (L/s/ha) 24.6 24.6 18.0 30.2 12.2 0 52.8 52.8 0 0 0 11.8 56.8 45.0 0
F.I.: first intake, R.P.: rough puddling, F.P.: finish puddling, S.W.M : shallow water management, M.S.I.: middle stage irrigation
Table 6 The four simulated cases

Case Water intake pattern Faucet opening
I 7th day in pattern 1 Fully open
II 7th day in pattern 1 Half open
III 13th day in pattern 2 Fully open
IV 13th day in pattern 2 Half open
table 5 refers to water intake for keeping paddy fields submerged after the periods of transplanting and
shallow water management. The two water intake patterns were determined as explained below on the
basis of the research on the characteristics of the irrigation water demand.
a) Finishing soil puddling is implemented on the day after rough soil puddling, and transplanting is
conducted in three days after soil puddling is finished.
b) In M.S.I., water intake is performed on the fifth day after transplanting and every fifth day after
that.
c) In the parcels for direct sowing culture, water intake is conducted for shallow water management
on the sixth day after the completion of the initial irrigation and once every six days after that.
d) The number of parcels where water intake is conducted on the first day of the initial irrigation
period is assumed to fall into two patterns: the maximum number of parcels available for
simultaneous irrigation (i.e., four parcels), and half the maximum number (i.e., two parcels)
In Water Intake Pattern 1 in table 5, the water demand becomes the largest on the 7th day when
shallow water management is implemented at four parcels. In Water Intake Pattern 2, the water
demand becomes the largest on the 13th day when shallow water management is implemented at two
parcels and middle stage irrigation is implemented at two parcels. Accordingly, the 7th and the13th
days were used for the simulations.
The aim of the water distribution simulations was to examine the feasibility of smooth water
distribution to the paddy fields used in this study. In light of this, it was necessary to determine which
parcels would be under relatively poor conditions for water intake so that water distribution to these
parcels could be simulated. Parcels No. 8, No. 9, No. 10 and No.11 in figure 8 were assumed to be
such parcels in the simulations.
Usually, farmers leave a faucet half open for irrigating their paddy fields. In view of this, two
patterns of faucet opening (i.e., fully open or half open) were used in the simulations.
Table 6 shows the four different combinations of water intake patterns and faucet opening
patterns (i.e., four cases) used in the simulations.
3.3 Simulation results and analysis

Simulation results are shown in figures 9 - 12. In Case I and Case II in which shallow water
management was conducted at four parcels, it took 13 hours and 15.5 hours, respectively, to complete
the water intake. In other words, more than 12 hours were necessary for the both parcels to complete
water intake for shallow water management. As mentioned above, farmers need to stop water intake
after visually confirming before dark that the entire surface of their parcels is submerged for shallow
water management. Thus, measures are necessary to reduce the time for completing water intake.
In Case III, water intake for middle-stage irrigation was completed in five hours. This
simulation result is acceptable to farmers in a cold region such as Hokkaido, because they are
encouraged to conduct water intake early in the morning or during the night to allow the temperature
of water in their paddy fields to rise during the day. In contrast, half a day was necessary to complete
water intake for shallow water management in Case III. In Case IV, it took more than half a day to
complete water intake for both middle-stage irrigation and shallow water management. The
researchers in this study investigated water demand characteristics in the paddy field zone, and
simulated the shallow water management that was identified in the investigation. The simulation
results clearly indicate that the quantity of water supplied in one water intake operation for shallow
water management is very large and that more than half a day is necessary to complete the water
intake even when simultaneous water intake is conducted at only two parcels.
A relatively large amount of water is necessary for soil puddling at parcels for transplanting
culture and shallow water management at parcels for direct sowing culture. If water is supplied to
paddy fields for these purposes on the same day, many hours are necessary to complete water intake.
In the period when soil puddling and shallow water management are intensively conducted, it seems
effective to reduce the amount of water intake for shallow water management at the parcels for direct
sowing culture by controlling the lowering of the groundwater level, because the culverts used for
supplying water to and discharging water from the parcels in the paddy field zone are equipped with
sluices that are available for groundwater level adjustment.
4. Conclusion
In this study, observation data on the groundwater level and other factors in large-scale paddy
fields to which subirrigation is applied were used for estimating the quantity of water intake necessary
in various farm operation periods and also for understanding the characteristics of irrigation water
demand in relation to transplanting culture and direct sowing in submerged fields. Water distribution
simulations were also done in which the water demand characteristics were taken into account, on the
assumption that improvement of paddy fields will be enhanced in the future toward increasing the
percentage of direct sowing in submerged fields.
The researchers in this study will look into the characteristics of water demand in relation to
other direct sowing methods that are expected to be applied to an increasingly large area of paddy
fields in the region used for the simulations. The results of the investigation on the water demand
characteristics will be used for the research on the management of irrigation water distribution. The
simulation results in this study will be explained to farmers and administrators of irrigation facilities
for discussion with them regarding feasible solutions to the problems of current water distribution
management.
References
HASEGAWA, K., NAKAMURA, K. and HIDESHIMA, Y., 2001, “Techniques for Calculating Water
Distribution in a Paddy Field Pipeline System and for Evaluating the Sufficiency Level of Water
Intake.” The 45th Workshop on Technology Research by the Hokkaido Regional Development
Bureau. Agriculture section No.8. (in Japanese)
Hokkaido Agricultural Development and Promotion Association., 2005, Agricultural Production
Techniques in Hokkaido (The 3rd edition), 8-13. (in Japanese)
NAITO, K., 1988, Pipeline - Design, Construction and Maintenance. The Agricultural Upland
Development Association, 131-172. (in Japanese)
SHIMURA, Y., MARUYAMA, T., CHO, T.,YAMAMOTO, T., SUZUKI, M., TABUCHI, T.,
KAIDA, Y., MITSUNO, T., SHIRAIWA, T. and SENGA, Y., 1987, New Study on Irrigation
and Drainage for Farming, Asakura Publishing Co., Ltd., 51.(in Japanese)
THE FUTURE
C-03
Development of a Simple Method of Discrimination between

the Dojo and Kara-dojo Loaches for the Conservation of
Japan’s Rural Ecosystem
Noriyuki Koizumi*, Kazuya Nishid**, Atsushi Mori*, Keiji Watabe*,

Takeshi Takemura*
* National Institute for Rural Engineering, 2-1-6 Kannondai, Ibaraki 305-8609, Japan
** National Research Institute of Far Seas Fisheries, 2-12-4 Fukuura, Kanagawa 236-8648,
Japan
1)
Corresponding author: koizumin@affrc.go.jp
ABSTRACT
We developed a simple method for the discrimination between two loaches, the Dojo, Misgurnus
anguillicaudatus and the Kara-dojo, Paramisgurnus dabryanus for helping in the conservation of
Japan’s rural ecosystem. The Dojo is a symbolic domestic loach, while the Kara-dojo is an alien fish
originally imported from China and Korea. In recent years the distribution of the Kara-dojo has
expanded nationwide, but it is difficult to even identify each loach, because the loaches have a similar
body shape and the existing methods have been established based on recondite taxonomy.
In this study, a simple discriminant function composed of two variables was statistically developed
using each of 312 Dojo and 72 Kara-dojo preliminarily identified by DNA analysis. By substituting
measurements of standard length, SL, caudal peduncle depth, CPD and the third barbell length, BL in
the function [Y = (24.1*CPD + 72.1*BL)/SL – 8.14], the calculated function score, Y allowed for the
correct classification of the Dojo (Y <= 0) and Kara-dojo (Y > 0) 97.7% and 88.5% of the time,
respectively (95.4% in total). This function could easily be used by beginners. We discuss the
features of the function and some limitations and considerations in its use.
Keywords: Endangered and alien fish species, discriminant function analysis, biodiversity,
agricultural canals and ditches
1. Introduction
The Dojo loach, Misgurnus anguillicaudatus (left in Fig.1) is widely distributed in East Asia.
This loach inhabits agricultural canals and ditches around rice paddies and also is used as local food in
some areas. Furthermore, in Japan the Dojo is considered as a kind of symbolic species that supports
rural ecosystem (Mizutani and Mori, 2009). However, in recent years the Kara-dojo loach,
Paramisgurnus dabryanus (right in Fig. 1) has been often observed along with the Dojo (Kanou,
2007; Koizumi et al., 2010; Shimizu and Takagi, 2010). The Kara-dojo is an alien species to Japan.
This species has been originally imported from China and Korea, to use not only for human
consumption but also feed for birds (Oliva and Hensel, 1961; Fujita, 2007). Perhaps in part the
imported fish have been artificially released and remained and then appear to have expanded their
distribution in rural areas of Japan.
10mm
Dojo loach, Misgurnus anguillicaudatus Kara-dojo loach, Paramisgurnus dabryanus
Figure 1. Photographs of the Dojo loach, Misgurnus anguillicaudatus (left) and the Kara-dojo loach,
Paramisgurnus dabryanus (right)
The existence of the Kara-dojo loach in Japan may cause some serious problems for
conservation soundness of the rural ecosystem (Fujita, 2007; Shimizu and Takagi, 2010). That is,
since body shape of the Kara-dojo is similar to that of the Dojo, it is difficult to visually discriminate
between both species. In addition both loach species appear to have common ecology, although the
details have not been sufficiently elucidated. Therefore it is possible that the Kara-dojo will perfectly
occupy the habitats of the Dojo in the future. Hybrids between both species might also be created
and such fish may induce disturbances of the genetic diversity. At present, with consideration for
these situations, the Dojo has been assigned as a species in the Red Data List of Japan (Ministry of the
Environment of Japan, 2013).
In this study we attempted to develop a discriminant function between the Dojo and Kara-dojo
loaches using measurement data of body parts such as body length and depth. Actually it is possible
to identify both loaches using categorical traits such as the number of scales and bones (Yang, et al.,
1994; Park et al., 2006) and DNA analysis (Koizumi et al., 2010; Shimizu and Takagi, 2010).
However, these methods ordinarily require relevant practical experiences and knowledge of recondite
taxonomy. Hence we performed a challenge to develop a simple method, so even beginners can use
it.
Standard length, SL
Head length, HL
Body depth, BD Caudal peduncle depth, CPD
Third barbell length of the right Caudal peduncle length, CPL

on the upper jaw, BL
Vertebral edge
Venter Right side
Figure 2. Six body parts measured in this study
2.1 Loach specimens

We examined a total of 312 and 72 specimens of the Dojo and Kara-dojo loaches, respectively.
Whether all specimens were the Dojo or Kara-dojo was previously identified using sequences of
chytochorome b gene in mitochondrial DNA (Koizumi et al., 2010). These specimens were collected
from agricultural canals and ditches at 45 sites in Tochigi Prefecture, July to September 2008. After
the collection, all specimens were immediately preserved in 99.5% ethanol and then stored at -30 °C in
the laboratory until the measurement of their body parts.
2.2 Measurement of body parts

A total of six body parts for each specimen, namely, standard length, SL, head length, HL, body
depth, BD, caudal peduncle depth and length, CPD and CPL, respectively and the third barbell length
of the right on the upper jaw, BL (Fig. 2) were measured in 0.1 mm with a digital caliper. These
body parts were chosen based on previous studies (Yang et al., 1994; Park et al., 2006; Shimizu and
Takagi, 2010) in consideration of ease of measurement for beginners.
2.3 Statistical analysis of measurement data

After the measurement, averages, standard deviations and ranges were estimated as fundamental
statistics for each body part of the Dojo and Kara-dojo loaches. And then we standardized values of
the measurement parts by dividing the part by SL, to observe obvious and relative differences between
both loaches. For example, the HL standardized by SL, HLSL% is denoted by eq. 1.
HLSL% = HL / SL × 100 (1)
Estimation of fundamental statistics, t-tests for differences of averages and examination of correlations
were performed for the standardized body parts.
A discriminant function was constructed using the data of the standardized body parts with the
software Ekuseru-Toukei 2010 (Social Survey Research Information Co., Ltd). Two species names
were used as external criteria, Y (0 and 1 of the Dojo and Kara-dojo, respectively), while the five
standardized parts were independent variables, Xi in this analysis. Variable selection using the
following stepwise method to remove multicollinearity of variables in the function and Mahalanobis'
generalized distance were adopted in the process of the construction. Statistical significance and
ratios of correlation and correct discrimination were examined as validity of the obtained function.
Table 1 Averages, standard deviations (s.d.) and ranges of measured six body parts of the Dojo and
Kara-dojo loaches
Dojo (n 1) = 312) Kara-dojo (n = 72)

Body parts
Average ± s.d. Range (min - max) Average ± s.d. Range (min - max)
Standard length, SL 60.0 ± 19.7 22.3 - 140.0 55.2 ± 14.0 29.0 - 91.0
Head length, HL 10.6 ± 2.9 5.1 - 22.4 10.2 ± 1.8 6.2 - 15.1
Body depth, BD 7.7 ± 2.9 2.1 - 20.9 8.7 ± 2.5 4.0 - 14.8
Caudal peduncle depth, CPD 6.1 ± 2.2 2.0 - 14.8 6.8 ± 2.3 2.7 - 12.4
Caudal peduncle length, CPL 8.3 ± 3.4 1.1 - 20.6 8.1 ± 2.7 3.7 - 16.1
The third barbell length, BL 3.4 ± 1.3 0.6 - 7.9 5.3 ± 1.3 3.1 - 8.2
Note) 1) the number of specimens
3. Results
3.1 Comparison of body parts

Table 1 shows fundamental statistics for six body parts of the Dojo and Kara-dojo loaches.
Averages of SL, HL, BD, CPD, CPL and BL for the Dojo were 60.0, 10.6, 7.7, 6.1, 8.3 and 3.4mm,
respectively, while those for the Kara-dojo were 55.2, 10.2, 8.7, 6.8, 8.1 and 5.3mm, respectively.
Since these data was not standardized, differences of averages between the loaches appear
indeterminate (Table 1).
Fig. 3 shows histograms for five body parts standardized by SL of the Dojo and Kara-dojo
loaches including fundamental statistics and results of t-test. Due to the standardization based on SL,
the histograms obviously displayed that the modes of BDSL%, CPDSL% and BLSL% for the Dojo were
smaller than those of the Kara-dojo. Averages of BDSL%, CPDSL% and BLSL% for the Dojo were 12.7,
10.0 and 5.6, respectively, while those for the Kara-dojo were 15.6 and 12.1 and 9.7, respectively (Fig.
3). Further differences of averages between both loaches were statistically significant for these
standardized parts (p < 0.01), although those of the other parts were also significant (p < 0.01).
HLSL% BDSL% CPDSL%

Dojo Kara-dojo Dojo Kara-dojo Dojo Kara-dojo
18.1 ± 1.7 18.9 ± 1.9 12.7 ± 1.4 15.6 ± 1.3 10.0 ± 1.0 12.1 ± 1.5
14.1 - 24.7 15.1 - 23.0 5.4 - 16.7 11.2 - 18.0 7.2 - 14.7 9.1 - 15.3
p < 0.01 p < 0.01 p < 0.01
100 20 100 40 120 20
)
)
Number of the kara-dojo specimens (n = 72,

50 10 50 20 60 10
Number of the Dojo specimens (n = 312,
0 0 0 0 0 0
14 18 22 25 5 9 13 18 7 10 13 16
CPLSL% BLSL%
Dojo Kara-dojo Dojo Kara-dojo Species
13.6 ± 2.4 14.4 ± 1.6 5.6 ± 1.2 9.7 ± 1.2 Average ± s.d.
4.3 - 19.3 11.4 - 18.9 1.8 - 9.0 6.7 - 12.6 Range (min – max)
p < 0.01 p < 0.01 Result of t-test
80 20 120
40 10 60 20
0 0 0 0
4 10 16 20 1 5 9 13
Figure 3. Histograms, fundamental statistics and results of t-tests for five body parts standardized by
standard length for the Dojo and Kara-dojo loaches
Fig. 4 shows correlation matrix among the standardized five body parts of the Dojo and
Kara-dojo loaches. The correlations among the parts differed depending on the pairs of chosen them.
The pairs of BDSL%-CPDSL% for both species significantly had the largest positive correlations (p <
0.01), showing 0.706 and 0.733 of correlation coefficients for the Dojo and Kara-dojo, respectively
(Fig. 4). On the other hand, the pairs of HLSL%-CPLSL% also significantly indicated the strongest
negative correlations with -0.419 and -0.479 of the coefficients for the Dojo and Kara-dojo,
respectively (p < 0.01).
Table 2 shows the discriminant function between the Dojo and Kara-dojo loaches using the
standardized parts including statistical significance and ratios of correlation and correct discrimination
in this analysis. The discriminant function was statistically significant (p < 0.01) and the calculated
discriminant scores, Y <= 0 indicate the Dojo, while Y > 0 is the Kara-dojo. The CPDSL% and BLSL%
were selected as two significant variables (p < 0.01) and the standardized coefficients of the variables
showed that BLSL% with 0.844 was more efficient factor than CPDSL% with 0.273 to decide the species
(Table 2). The correlation ratio was also high and the ratios of correct discrimination for the total,
Dojo and Kara-dojo were 95.3, 95.5 and 94.4, respectively.
HLSL% Upper: correlation coefficient of the Dojo (n = 312)

Lower: correlation coefficient of the Kara-dojo (n = 72)
**: p < 0.01
-0.011 Line: regression line
BDSL%
-0.094
BDSL%
-0.063 0.706**
CPDSL%
-0.376** 0.733**
CPDSL%
-0.419** 0.166** 0.222**

CPLSL%
-0.479** 0.254** 0.391**
CPLSL%
0.199** 0.503** 0.416** -0.015

BLSL%
0.358** 0.103** -0.050 -0.194
Figure 4. Correlation matrix among the five parts standardized by standard length of the Dojo and
Kara-dojo loaches
Table 2 Discriminant function between Dojo and Kara-dojo loaches using standardized measurement
parts in this study
Discriminant function Y Significant Correlation

Correct discrimination %
Y <= 0: the Dojo, Y > 0: the Kara-dojo probability ratio
Total: 95.3 (366 2) / 384 3))

Y = 0.241 CPDSL% ** + 0.721 BLSL% ** - 8.140
＜0.01 0.804 Dojo: 95.5 (298 / 312)
(0.273) 1) (0.884)
Kara-dojo: 94.4 (68 / 72)
1) 2) 3)
Note) ** p < 0.01, standardized coefficient, the number of correct discrimination, the total number of
specimens
4. Discussion
4.1 Features of the discriminant function

The discriminant function in this study (Table 2) allowed for the correct discrimination of the
Dojo and Kara-dojo loaches 95.5% and 94.4% of the time, respectively (95.3% in total). Such
discriminant functions with high accuracy were developed in some previous studies (Matsumiya et al.,
1984; Yu et al., 2010; Uglem et al., 2011). The silver carp Hypophthalmichthys molitrix and bighead
carp Aristichthys nobilis were classified 94.2% and 98.0% of the time, respectively (Yu et al., 2010),
while the wild and farmed Atlantic cod Gadus morhua were 100% and 95%, respectively (Uglem et
al., 2011). Although classification success depends on quality and accuracy of measurement data for
morphological characters, discriminant function will be a powerful statistical method to help with
difficult visual classification.
The CPDSL% and LBSL% were finally adopted as the significant variables of the discriminant
function in this study (Table 2). Both the CPDSL% and BLSL% of the Dojo loach were significantly
smaller than those of the Kara-dojo loach (Fig. 3). The similar observations for the CPD and BL
between the Dojo and Kara-dojo were reported in previous studies (Oliva and Hensel, 1961; Vasil’eva,
2001; Park et al., 2006; Kanou et al., 2007; Shimizu and Takagi, 2010). The CPD and BL appears a
kind of key indicators to distinguish between the Dojo and Kara-dojo.
4.2 Use of the discriminant function

The discriminant function in this study is simple and can be easily used even by beginners
(Table 2). Hence, this function would be expected to be applied to many specimens of the Dojo and
Kara-dojo loaches in other fields. Here, we should show that there are some limitations and
considerations for the function and its use, because the function is also a statistical model including
some errors.
First, it would be better to use specimens preserved in ethanol over a month, with consideration
for shrink of body parts during the preservation (Fox, 1996; Cunningham et al., 2000). And also this
function would be directly applied to specimens even if they are alive. We confirmed that the
function was effective using over 50 of such live specimens.
Second, measurement data of each part of loach specimens should be within the ranges in Table
1. The correct and incorrect discrimination using the data outside the ranges was not examined in
this study.
Third, if the calculated function score ranges -0.467 to 1.217, the loach species indicated by the
function might be incorrect. As shown the histogram of the discriminant scores in Fig.5, the function
incorrectly discriminated 14 of the 312 Dojo as the Kara-dojo and 4 of the 72 Kara-dojo as the Dojo.
Therefore careful attention and reinvestigation using taxonomic method and DNA analysis might be
needed for such specimens.
30 Correct discrimination (number of specimens)

Dojo -> Dojo (298)
Kara-dojo -> Kara-dojo (68)
Number of specimens
Incorrect discrimination
20
Dojo -> Kara-dojo (14)
Kara-dojo -> Dojo (4)
10 Range of incorrect discrimination

(-0.467 - 1.217)
0
-5 -4 -3 -2 -1 0 1 2 3 4
Score of discriminant function, Y
Figure 5. Histogram of scores of discriminant function for correct and incorrect discrimination
Acknowledgments
Our thanks go to Dr. Masakazu Mizutani and Mr. Gen-ichi Nakakuki for their collection of the
loach specimens, Mr. Park Myeong Soo and Mses. Ponthip Goto, Kyoko Yamanoi and Aiko Suzuki
for their supports of the laboratory experiment and the data analysis, and also Mr. Philip Johnson for
his English correction of this manuscript. This study was supported in part by a Grant-in-Aid for
Scientific Research (C-23580340 and C-25450367) from the Japan Society for the Promotion of
Science.
References
Cunningham, M.K., Granberry, W.F. and Pope, K.L., 2000, “Shrinkage of inland silverside larvae
preserved in ethanol and formalin.” North American Journal of Fisheries Management, 20,
816-818.
Fox, C.J., 1996, “Length changes in herring (Clupea harengus) larvae: effects of capture and storage
in formaldehyde and alcohol.” Journal of Plankton Research, 18, 483-493.
Fujita, A., 2007, “Scientific name for the loach “Kara-Dojo” distributed in Japan.” Japanese Journal
of Ichthyology, 54, 243-244. (in Japanese)
Kanou, K., Saito, S., Fuchigami, S., Imamura, A., Imai, H. And Taki, Y., 2007, “Occurrence patterns
and food habits of introduced alien loach Paramisgurnus dabryanus and native loach Misgurnus
anguillicaudatus at irrigation drainages around rice fields in the Watarase River System, Central
Honshu, Japan.” Aquaculture Science, 55, 109-114. (in Japanese with English abstract)
Koizumi, N., Mori, A., Nakakuki, G., Mizutani, M., Nishida, K., Takemura, T., Watabe, K. and Park,
M., 2010, “Elucidation of genetic clades for loach in Tochigi Prefecture.” Proceedings of the
Annual Conference 2010 of the Japanese Society of Irrigation, Drainage and Rural Engineering,
860-861. (in Japanese)
Matsumiya, Y., Kanamaru, H., Oka, M. and Tateishi, M.,1984, “Morphometric differentiation
between artificially-released red sea bream and 0-age wild fish by discriminant function.”
Nippon Suisan Gakkaishi, 50, 1179-1185. (in Japanese with English abstract)
Ministry of the Environment of Japan, 2013, The 4th Version of the Japanese Red Lists,
http://www.env.go.jp/press/press.php?serial=16264, February 2013.
Mizutani M. and Mori A. (ed.),2009, Conserving Habitat of Freshwater Fishes Inhabiting
Haruno-ogawa, Irrigation/Drainage Ditches, in Rice Paddies. Gakuhosya, 1-190. (in Japanese)
Oliva, O. and Hensel, K., 1961, “Some remarks on eastern Asiatic loaches of the genus Misgurnus
(Cobitidae).” Japanese Journal of Ichthyology, 8, 86-91.
Park, I.S., Nam, Y.K. and Kim, D.S., 2006, “Growth performance traits and gonad development of
induced reciprocal diploid and triploid hybrids between the mud loach (Misgurnus mizolepis
Gunther) and cyprinid loach (Misgurnus anguillicaudatus Cantor).” Aquaculture Research, 37,
1246-1253.
Shimizu, T. and Takagi, M., 2010, “Two genetic clades in populations of Paramisgurnus dabryanus,
an exotic invader in Ehime Prefecture, Japan.” Japanese Journal of Ichthyology, 57, 125-134.
(in Japanese with English abstract)
Uglem, I., Berg, M., Varne, R., Nilsen, R., Mork, J. and Bjørn, P.A., 2011, “Discrimination of wild
and farmed Atlantic cod (Gadus morhua) based on morphology and scale-circuli pattern.”
ICES Journal of Marine Science, 68, 1928-1936.
Vasil’eva, E.D., 2001, “Loaches (genus Misgurnus, Cobitidae) of Russian Asia. I. the species
composition in waters of Russia (with a description of a new species) and some nomenclature
and taxonomic problems of related forms from adjacent countries.” Journal of Ichthyology, 41,
553-563.
Yang, S.Y., Yang, H.J., Jeon, S.R., Nam, M.M., Min, M.S. and Kim, J.H., 1994, “Systematic study on
the fishes of the family Cobitidae (Pisces, Cypriniformes). 3. Taxonomic study on morphological
variation of the Misgurnus anguillicaudatus and M. mizolepis from Korea.” Bulletin of the
Institute for Basic Science, Inha University, 15, 79-86. (in Korean with English abstract)
Yu, H.X., Tang, W.Q. and Li, S.F., 2010, “Morphological changes of silver and bighead carp in the
Yangtze River over the past 50 years.” Zoological Research, 31, 169-176.
THE FUTURE
C-07
Determining Optimal Soil Moisture for Irrigated Rice in

Indonesia with System of Rice Intensification
Chusnul Arif*, Masaru Mizoguchi**, Budi Indra Setiawan*, Tsugihiro Watanabe***

*Department of Civil and Environmental Engineering, Bogor Agricultural University (IPB),
Bogor, Indonesia Email: chusnulipb@gmail.com, chusnul_arif@ipb.ac.id
**Department of Global Agricultural Sciences, The University of Tokyo, Japan
***Research Institute for Humanity and Nature, Kyoto, Japan.
In this study, an optimal combination of soil moisture for irrigated rice in Indonesia that maximizes
both yield and water productivity of system of rice intensification (SRI) paddy field was determined by
genetic algorithm (GA) model-based optimization. Before performing optimization, a formula to
describe yield by soil moisture and meteorological parameters was identified using multiple
non-linear regression analysis. The GA model was performed based on the identification process
according to the empirical data during three cropping seasons. Here, we classified soil moisture level
into three levels i.e. wet (W), medium (M) or dry (D) based on the soil water retention curve. As the
results, the optimal soil moisture was a combination of wet, wet, medium, and dry levels for initial,
crop development, mid-season and late season growth stages, respectively. We called this regime as
W-W-M-D regime. The wet level in the initial and crop development growth stages should be achieved
providing enough water for the plant to develop root, stem and tiller, and then the field can be drained
into the medium level with the irrigation threshold of field capacity to avoid spikelet sterility in
mid-season stage and finally, let the field in the dry level to save more water in the late season stage
when plant water requirement is minimum. By this scenario, it was simulated that the yield can be
increased up to 8.35% and water productivity up to 13.49% with saving water up to 12.28% compared
to the empirical data.
Keywords: system of rice intensification (SRI), non-flooded irrigation, water productivity, water
saving, genetic algorithm
1. Introduction
Recently, the scarcities of water resources and competition for their use have made water saving
the main challenge in maintaining the sustainability of rice farming. Therefore, water saving
technology becomes one of the priorities in rice research (Barker et al., 2000). Rice is highly possible
produced under water saving technology with system of rice intensification (SRI) in which continuous
flooded irrigation is not essential anymore to gain high yield and biomass production (Lin et al., 2011;
Sato et al., 2011; Zhao et al., 2011).
SRI is well-known as a set crop management practices for raising the productivity of irrigated rice
by changing the management of plants, soil, water and nutrients. Although some critics were
dismissed to the SRI (Dobermann, 2004; Sheehy et al., 2004; Sinclair and Cassman, 2004), however,
its benefits have been validated in 42 countries of Asia, Africa and Latin America (Uphoff et al., 2011).
In the SRI paddy field, non-flooding irrigation is applied in which the field is allowed dry during
particular time instead of keeping them continuously flooded, a practice called alternate wetting and
drying irrigation (Van der Hoek et al., 2001).
Many experiments have been conducted by comparing continuous flooding and non-flooding
regimes under SRI (Barison and Uphoff, 2011; Chapagain and Yamaji, 2010; Choi et al., 2012;
Hameed et al., 2011; Sato et al., 2011; Zhao et al., 2011). Water productivity can be raised and water
can be saved significantly, as reported in studies that provide data for different countries, e.g., 28% in
Japan (Chapagain and Yamaji, 2010), 40% in Eastern Indonesia (Sato et al., 2011), and 38.5% in Iraq
(Hameed et al., 2011). Also by SRI, the land productivity raised more than double in Madagascar
(Barison and Uphoff, 2011), 78% in Eastern Indonesia (Sato et al., 2011), 65% in Afghanistan
(Thomas and Ramzi, 2011), 42% in Iraq (Hameed et al., 2011), and 11.3% in China (Lin et al., 2011).
However, the optimal wet and dry levels (represented by soil moisture) in each growth stage is still
unclear because there is lack information study on optimizing water management of SRI paddy field.
Thus, the current study was undertaken to find optimal soil moisture level in each growth stage to
maximize both yield and water productivity during cultivation period.
In the irrigation planning model, there are many factors to be considered, such as crop water
requirement, production function, precipitation, soil water balance including irrigation water, plant
growth stage, etc (Zhang et al., 2008). It is difficult problem to find the optimal or near optimal
solution with traditional optimization methods because the limitations in integrating of multi-factors in
the model. Thus, genetic algorithm (GA) proposes global optimization search with many remarkable
characteristics by searching the entire population instead of moving from one point to the next as the
traditional methods (Kuo et al., 2000).
GA has the ability to rapidly search a global optimal value of a complex objective function using a
multi-point search procedure involving crossover and mutation processes (Goldberg, 1989). GA
differs from the traditional optimization and other search methods in the following ways: (1) GA
works with a coding of the parameter set, not the parameters themselves, (2) GA searches from
population of points, not a single point, (3) GA uses objective function, not derivatives or other
auxiliary knowledge, and (4) GA uses probabilistic transition rules, not deterministic rules (Goldberg,
1989). GA has been applied to several irrigation planning applications (Kuo et al., 2000; Raju and
Kumar, 2004; Wardlaw and Bhaktikul, 2004; Zhang et al., 2008). However, optimizing water
management in any SRI paddy fields have not yet been achieved by finding the optimal soil moisture
in each growth stage.
Therefore, the main objectives of this study was to find the optimal water management by
determining optimal combination of soil moisture levels using GA model in maximizing both yield
and water productivity.
2.1 Field Experiments

The optimization process was carried out based on the field experiments in the experimental
paddy field in the Nusantara Organics SRI Center (NOSC), Sukabumi, West Java, Indonesia located at
06º50’43”S and 106º48’20”E, at an altitude of 536 m above mean sea level (Fig. 1) during three
cropping seasons (Table 1).
Source: earth.google.com (2012)
Fig. 1 Experimental field location in West Java, Indonesia.
There were four plots and each plot was planted with the variety of rice (Oryza sativa L), Sintanur
using the following SRI elements: single planting of young seedlings spaced at 30 cm × 30 cm,
applying an organic fertilizer at 1 kg/m2 in the land preparation, but no chemical fertilizer. The
weeding was performed every 10 days in the period between 10 and 40 days after transplantation
supplying local indigenous microorganism to enhance biological activity in the soils (Uphoff et al.,
2011).
Table 1 Cultivation period of each cropping season

Period Planting date Harvesting date Season
First 14 October 2010 8 February 2011 Rainy
Second 20 August 2011 15 December 2011 Dry - Rainy
Third 22 March 2012 5 July 2012 Rainy - Dry
Each plot was irrigated under non-flooded condition with different soil moisture level in each
growth stage. Here, during cultivation period, growth stage was divided into four stages, i.e., initial,
crop development, mid-season and late season stages (Allen et al., 1998; Mohan and Arumugam,
1994; Tyagi et al., 2000; Vu et al., 2005). Also, soil moisture level was classified into three levels i.e.
wet (W), medium (M) or dry (D) based on the soil water retention curve as presented in Fig. 2. The
wet level was achieved when pF value was between 0 and 1.6 which was the air entry value for this
soil. The medium level was achieved when pF value was between 1.6 and 2.54 which was the field
capacity value. When the soil was drier than the medium level, the condition was regarded as the dry
level.
5.0
4.0
Dry
3.0
Medium
pF
2.0
Wet
1.0
Saturated
0.0
0.0 1.0
Volumetric water content (cm3/cm3)
Fig. 2 Classification of soil moisture level during cultivation period.
Soil moisture was measured by 5TE sensor by Decagon Devices, Inc., USA. Meanwhile,
precipitation, solar radiation and air temperature were measured by Davis Vantage Pro2 weather
station. After harvesting, yield in each plot was obtained to determine water productivity with respect
of total water input (Bouman et al., 2005) by the following equation:
Y
WP = (1)
∑ (I + P)
where Y is yield (ton/ha), I is total irrigation (mm), P is precipitation (mm) and WP is water
productivity (g grain/kg water).
2.2 Modeling approach
2.2.1 Identification procedure

Since we focused on water management, all of inputs of production such as fertilizer and seeds
were given at same levels except for water input. Therefore, identification process was carried out to
correlate between soil moisture and weather parameters as the inputs with yield as the outputs before
performing the GA model by the following equation:
Y = f(SM1, SM2, SM3, SM4, Rs, T) (2)
where, Y is yield (ton/ha), Rs is total solar radiation (MJ/m2/season), T is average air temperature
o
( C), SM1, SM2, SM3, SM4 are the average soil moisture for initial, crop development, mid-season,
and late season stages (cm3/cm3). Since there is no mathematical equation from previous research
because of the complexity of this relationship, we implemented neural networks model to show its
correlation because neural networks model deals with complex system such as in agricultural system
(Hashimoto, 1997). The model consisted three layers, i.e. input, hidden and output layers as presented
in Fig.3.
Fig. 3 Structure of neural networks model to estimate yield based on environmental parameters.
2.2.2 Optimization procedure

Optimization process was carried out by the GA model by the following objective function:
F = a 1 Y + b 2 WP (3)
Maximize F, subject to:
SMmin ≤ SM1, SM2, SM3, SM4 ≤ SMmax (4)
where, a1 and b1 are weights for yield and water productivity and their values are 0.5 and 0.5,
respectively. SMmin and SMmax are the minimum and maximum soil moisture levels from the
empirical data during three cropping seasons (cm3/cm3). Since both yield and water productivity have
different units, their values were normalized using the maximum and minimum values based on
empirical data.
3.1 Meteorological conditions

Table 2 summarizes the climatic data during the experiments in three seasons. There are two
seasons in Indonesia classified based on the pattern of precipitation. Here, precipitation among seasons
was quite different in which the highest intensity occurred in the first season with total precipitation of
1332 mm in rainy season. Consequently, different pattern in precipitation corresponded to the different
pattern of solar radiation. The lowest solar radiation occurred in the first season with total values of
1464 MJ/m2/season. Meanwhile, temperatures among the seasons were quite same in which maximum
temperature was 32.8oC and minimum temperature was 16.2oC in the second season.
Table 2 Meteorological data during experiments
Temperature (oC)
Precipitation Solar radiation
Seasons (mm) Minimum Average Maximum (MJ/m2/season)
I (Rainy) 1332 19.5 23.5 31.9 1464
II (Dry-Rainy) 626 16.2 24.0 32.8 1827
III (Rainy-Dry) 551 17.4 24.3 32.3 1829
3.2 Correlation between yield and soil moistur levels

Fig. 4 shows model validation results between observed and estimated yield by neural networks
model. The model estimated yield with high correlation to the observed yield (R2 = 0.93) which
indicated that yield is mainly affected by soil moisture levels and weather conditions when fertilizer
and others inputs were given at same level in all plots.
12
Observed Yield (ton/ha)
10
y = 1.00x
8 R² = 0.93
2
2 4 6 8 10 12
Yield model (ton/ha)
Fig. 4 Model validation of neural networks model to estimate yield
Fig. 5 shows linear correlation between the average soil moisture in each growth stage and the
yield. The third season obtained the highest yield compared to other seasons. The average yield was
4.77, 4.23 and 9.38 ton/ha for the first, second and third seasons, respectively. Hence, soil moisture
levels have correlation to the yield for all growth stages.
In the initial and crop development stages, soil moisture had positive correlation to yield with an
R2 of higher than 0.6. This result revealed that at higher soil moisture levels, more yield was produced.
In the initial stage, the maximum yield was produced when the soil moisture level was over the
saturation border indicating shallow standing water was occurred in the field. Then, in the crop
development stage, the maximum yield was achieved when the soil moisture level was close to the
saturation border. The field condition in the crop development stage was drier than that in the initial
stage, even though both conditions were classified as wet condition.
On the contrary, soil moisture had negative correlation to yield in the mid-season and late season
stages. Based on the empirical data, the mid-season stage was probably the transition in which the
water can be drained to produce more yields. Here, the maximum yield was obtained when the soil
moisture level was higher than that the field capacity border. This revealed that the medium level was
appropriate to produce more yields by draining water in the mid-season stage. Then, in the late season
stage, the driest condition can be applied to save more water without a loss of significant yield.
0.65 Initial 0.65 Crop development
y = 0.01x + 0.51
0.63 0.63
Soil moisture (cm3/cm3)
R² = 0.65

Saturated border Saturated border
0.60 0.60
y = 0.01x + 0.49
0.58 R² = 0.82 0.58
Season 1 Season 1
0.55 0.55
Season 2 Season 2
0.53 Season 3 0.53
Season 3
0.50 0.50
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
Yield (ton/ha) Yield (ton/ha)
0.65 0.60
Mid-season Late season
0.60 y = -0.01x + 0.58 y = -0.02x + 0.61
0.55
R² = 0.44

R² = 0.27
0.55 0.50
0.50 0.45
Season 1 Field capacity border Season 1

0.45 0.40
Field capacity border Season 2 Season 2
0.40 0.35
Season 3 Season 3
0.35 0.30
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
Yield (ton/ha) Yield (ton/ha)
Fig. 5 Linear correlation between the average soil moisture and the yield in each growth stage
3.3 Optimal soil moisture by the GA model

As previously mentioned, during three cropping seasons the highest yield was obtained in the
third season. So, we used meteorological data in this season as the inputs, and then the GA model
searched the optimal soil moisture in each growth stage
Fig. 6 shows the evolution curves of fitness values between their maximum, average and
minimum values in each generation. All values increased sharply from the first to the tenth generation,
and then increased gradually until the 38th generation. After the 38th generation, the all fitness values
were convergent until the end of generation and their values were 0.28. This means that the global
maximum value was obtained because all of their maximum, average and minimum values were the
same.
0.30
0.25
Fitness
0.20
Max fitness
0.15 Ave fitness
0.10
0 10 20 30 40 50 60 70 80 90 100
Generation numbers
Fig. 6 Evolution curves in searching for a maximal value of fitness function
Fig. 7 shows the evolution curves of soil moisture level in each growth stage in obtaining fitness
values in Fig. 6. SM1 and SM2 converged faster than the other growth stages; their values reached the
asymptote before the tenth generation. Meanwhile, SM3 became convergent most slowly; in the 38th
generation, at which the fitness value was also starting to be convergent. This means that the optimal
soil moisture level in each growth stage that maximizes the yield and water productivity was obtained
from the model simulation based on the GA procedure after the 38th generation.
0.700
Wet
0.600
Medium
0.500
0.400
Dry
0.300
0 10 20 30 40 50 60 70 80 90 100
Generation numbers
SM1 SM2 SM3 SM4
Fig. 7 Evolution curves in searching the optimal values of soil moisture in each growth stage
Table 3 shows the optimal soil moisture level in each growth stage obtained by the GA model.
Four irrigation regimes with the combinations of soil moisture levels from the field measurements in
the third season are also represented in the table as the comparison. The optimal combination of soil
moisture levels in the growth stages obtained in this chapter was 0.622 (wet), 0.593 (wet), 0.455
(medium), and 0.350 cm3/cm3 (dry) for SM1, SM2, SM3 and SM4, respectively. We called this regime
as W-W-M-D. By this scenario, it was simulated that the yield can be increased up to 8.35% and water
productivity up to 13.49% with water saving up to 12.28%.
From this simulation, it was shown that during the first and second stages keeping the field in the
wet level is important to fulfill the plant water requirement for vegetative development. This result
was supported by the empirical data that the maximum yield was obtained when a wet level was
developed in the field. In SRI paddy field, to avoid continuous flooding is one of the main elements
because rice plants cannot grow best under limited oxygen in the soil, thus plants should be given just
enough water at saturated condition to meet their requirement for root, stem and tiller development
(Uphoff et al., 2011).
Then, the field can be drained into the medium level in the third stage when the plants focusing
on the reproductive stage (flowering and panicle development). The medium level is important in
developing aerobic condition to avoid spikelet sterility particularly around the flowering time
(Bouman et al., 2005). Finally, the field should be drained into dry level in the last stage when plant
water requirement is minimum to save water as reported in the previous studies (Doorenbos and
Kassam, 1979; Uphoff et al., 2011; Zawawi et al., 2010). This recommendation was also supported by
the empirical data that medium and dry levels in the mid-season and late season stages resulted in the
maximum yield.
Table 3 Optimal soil moisture level in each growth stage and its comparison to the irrigation
regimes in the third season
Irrigation regimes GA model
Optimal
Components Plot 1 Plot 2 Plot 3 Plot 4 Regime Level
3 3
Soil moisture (cm /cm )
Initial (SM1) 0.622 0.602 0.611 0.586 0.622 Wet
Crop development (SM2) 0.592 0.585 0.593 0.563 0.593 Wet
Mid-season (SM3) 0.522 0.488 0.472 0.455 0.455 Medium
Late season (SM4) 0.505 0.401 0.456 0.350 0.350 Dry
Yield (ton/ha) 10.00 9.38 8.75 9.38 10.84
Total irrigation (mm) 343 295 305 272 301
Total precipitation (mm) 551 551 551 551 551
Water productivity (g grain/ kg
water) 1.12 1.11 1.02 1.14 1.27
Water saving (%) - 13.86% 11.01% 20.65% 12.28%
4. Conclusions
The optimal combination of soil moisture levels was estimated by the GA model for the growth
stages to maximize both the yield and the water productivity of the SRI paddy field. The simulation
was performed based on the identification process using the empirical data during the three cropping
seasons. As a result of the simulation, the optimal values were estimated at 0.622 (wet), 0.593 (wet),
0.455 (medium), and 0.350 cm3/cm3 (dry) for the initial, crop development, mid-season, and late
season growth stages, respectively. We called this regime as W-W-M-D regime. The wet level in the
initial and crop development growth stages should be achieved to provide enough water for vegetative
development, and then the field can be drained with the irrigation threshold of field capacity to avoid
spikelet sterility in the mid-season stage and finally, to complete the production, it is important to let
the field dry to save more water in the late season stage. By this scenario, it was estimated that the
yield can be increased up to 8.35% and water productivity up to 13.49% with water saving up to
12.28%.
Acknowledgments
We are grateful to the Directorate of Higher Education, Ministry of National Education, Republic
of Indonesia for generous financial support through grant of International Research Collaboration and
Scientific Publication. Also, the study was partially supported by GRENE (Green Network of
Excellence) project of MEXT in Japan and sponsored by Research Institute of Humanity and Nature
(RIHN), Japan collaborated with Bogor Agricultural University (IPB) launched in 2011 entitled
Designing Local Framework for Integrated Water Resources Management.
References
Allen, R.G., Pareira, L.S., Raes, D., Smith, M., 1998, Crop Evapotranspiration Guidelines for
computing crop water requirements. FAO - Food and Agriculture Organization of the United
Nations, Rome.
Barison, J., Uphoff, N., 2011, Rice yield and its relation to root growth and nutrient-use efficiency
under SRI and conventional cultivation: an evaluation in Madagascar. Paddy Water Environ 9,
65-78.
Barker, R., Dawe, D., Tuong, T.P., Bhuiyan, S.I., Guerra, L.C., 2000, The outlook for water resources
in the year 2020: challenges for research on water management in rice production. International
Rice Commission Newsletter 49, 7-21.
Bouman, B.A.M., S.Peng., Castaneda, A.R., Visperas, R.M., 2005, Yield and water use of irrigated
tropical aerobic rice systems. Agr Water Manage 74, 87-105.
Chapagain, T., Yamaji, E., 2010, The effects of irrigation method, age of seedling and spacing on crop
performance, productivity and water-wise rice production in Japan. Paddy Water Environ 8,
81-90.
Choi, J.D., Park, W.J., Park, K.W., Lim, K.J., 2012, Feasibility of SRI methods for reduction of
irrigation and NPS pollution in Korea. Paddy Water Environ published online by Springerlink
Feb. 9.
Dobermann, A., 2004, A critical assessment of the system of rice intensification (SRI). Agr Syst 79,
261-281.
Doorenbos, J., Kassam, A.H., 1979, Yield response to water. FAO Irrigation and Drainage Paper 33.
FAO, Rome.
Goldberg, D.E., 1989, Genetic algorithms in search optimization and machine learning.
Addison-Wesley, Reading, Massachusetts.
Hameed, K.A., Mosa, A.K.J., Jaber, F.A., 2011, Irrigation water reduction using System of Rice
Intensification compared with conventional cultivation methods in Iraq. Paddy Water Environ 9,
121-127.
Hashimoto, Y., 1997, Applications of artificial neural networks and genetic algorithms to agricultural
systems. Comput Electron Agr 18, 71-72.
Kuo, S.F., Merkley, G.P., Liu, C.W., 2000, Decision support for irrigation project planning using a
genetic algorithm. Agr Water Manage 45, 243-266.
Lin, X.Q., Zhu, D.F., Lin, X.J., 2011, Effects of water management and organic fertilization with SRI
crop practices on hybrid rice performance and rhizosphere dynamics. Paddy Water Environ 9,
33-39.
Mohan, S., Arumugam, N., 1994, Irrigation crop coefficient for lowland rice. Irrigation and Drainage
Systems 8, 159-176.
Raju, K.S., Kumar, D.N., 2004, Irrigation planning using Genetic Algorithms. Water Resour Manag
18, 163-176.
Sato, S., Yamaji, E., Kuroda, T., 2011, Strategies and engineering adaptions to disseminate SRI
methods in large-scale irrigation systems in Eastern Indonesia. Paddy Water Environ 9, 79-88.
Sheehy, J.E., Peng, S., Dobermann, A., Mitchell, P.L., Ferrer, A., Yang, J.C., Zou, Y.B., Zhong, X.H.,
Huang, J.L., 2004, Fantastic yields in the system of rice intensification: fact or fallacy? Field
Crop Res 88, 1-8.
Sinclair, T.R., Cassman, K.G., 2004, Agronomic UFOs. Field Crop Res 88, 9-10.
Thomas, V., Ramzi, A.M., 2011, SRI contributions to rice production dealing with water management
constraints in northeastern Afghanistan. Paddy Water Environ 9, 101-109.
Tyagi, N.K., Sharma, D.K., Luthra, S.K., 2000, Determination of evapotranspiration and crop
coefficients of rice and sunflower with lysimeter. Agr Water Manage 45, 41-54.
Uphoff, N., Kassam, A., Harwood, R., 2011, SRI as a methodology for raising crop and water
productivity: productive adaptations in rice agronomy and irrigation water management. Paddy
Water Environ 9, 3-11.
Van der Hoek, W., Sakthivadivel, R., Renshaw, M., Silver, J.B., Birley, M.H., Konradsen, F., 2001,
Alternate wet/dry irrigation in rice cultivation: a pratical way to save water and control malaria
and Japanese encephalitis?, Research Report 47. International Water Management Institute,
Colombo, Sri Lanka.
Vu, S.H., Watanabe, H., Takagi, K., 2005, Application of FAO-56 for evaluating evapotranspiration in
simulation of pollutant runoff from paddy rice field in Japan. Agr Water Manage 76, 195-210.
Wardlaw, R., Bhaktikul, K., 2004, Application of genetic algorithms for irrigation water scheduling.
Irrig Drain 53, 397-414.
Zawawi, M.A.M., Mustapha, S., Puasa, Z., 2010, Determination of water requirement in a paddy field
at seberang perak rice cultivation area. Journal - The institution of Engineers 71, 32-41.
Zhang, B., Yuan, S.Q., Zhang, J.S., Li, H., 2008, Study of corn optimization irrigation model by
genetic algorithms, IFIP International Federation for Information Processing. Springer, pp.
121-132.
Zhao, L.M., Wu, L.H., Wu, M.Y., Li, Y.S., 2011, Nutrient uptake and water use efficiency as affected
by modified rice cultivation methods with reduced irrigation. Paddy Water Environ 9, 25-32.
Session 4
THE FUTURE
D-01
Decrease of Egg-masses for the Japanese Brown Frog (Rana ja-

ponica) after Land Consolidation Project in Paddy Field Area,
Japan
Keiji Watabe*1), Atsushi Mori*, Noriyuki Koizumi*, Takeshi Takemura*

and Kazuya Nishida**
* Institute for Rural Engineering, National Agriculture and Food Research Organization,
JAPAN
** National Research Institute of Far Seas Fisheries, Fisheries Research Agency
1)
kwatabe@affrc.go.jp
Paddy fields provide not only our major crops but also habitats of small animals such as frogs and
insects that play an important role for sustaining a sound ecosystem. Land consolidation projects of
Japan contribute to productivity advance in the crops, while the projects sometimes damage small
animal habitats. Toward their habitat conservation along with sustainable and productive agriculture,
it is necessary to observe population of the small animals related to process of the projects. In this
study, we investigated the number of egg-masses of the endangered Japanese Brown Frog (Rana ja-
ponica) before and after a land consolidation project.
Survey sites consisted of the inside and outside project area, IPA and OPA, respectively, were located
in rice paddy fields in the Kanto region, Japan. In IPA, cutting and banking soil along with construct-
ing concrete canals was executed in 2010. The number of brown frog egg-masses, which indicates
abundance of mature females, was counted with walking along all paddy levees of IPA and OPA in
2008-2013. Average size of the egg-mass density (number / km) in IPA was 66 before the project
(2008-2009), while it decreased to only 7 after the project (2011-2013). Such a remarkable size
change was not observed in OPA. These results suggest that the brown frogs as well as the egg-masses
would diminish due to the project.
Keywords: Amphibia, rice paddy ecosystem, biodiversity, monsoon Asia, agricultural infrastructure
improvement
1. Introduction
Paddy fields provide not only our major crops but also habitats of small animals such as frogs
and insects that play an important role for sustaining a sound ecosystem. Land consolidation projects
in Japan include land readjustments, construction of irrigation and drainage facilities, farm road im-
provements, and contribute to productivity advances in agriculture. However, the projects sometimes
damage small animal habitats. Various small animals that were common species in paddy fields until
recent decades are now designated as endangered species in the national or prefectural red data books.
Taward land consolidation projects in environmentally-sensitive ways that achieve their habitat
conservation along with sustainable and productive agriculture, it is firstly necessary to observe
population of the small animals related to process of the projects.
This study focused on the Japanese Brown Frog, Rana japonica. The brown frogs are endemic
to Japan, however have been designated as an endangered species in some areas. Such projects are
thought to be one of the reasons of frog disappearance (e.g. Hasegawa 1995). Hence, we investigated
the number of their egg-masses before and after a land consolidation project.
2. Materials and methods
2.1 Study area

The study area was located in rice paddy fields in the Kanto Plain, Japan (Fig. 1; 36N, 140E;
altitude: 100m). The area consisted of the inside project area (“IPA”, 70,000m2) and outside project
area (“OPA”, 30,000m2).
In IPA, There were 37 small irregular-shaped paddy lots (average 1,561m2 / lot) with earthen
ditches until the summer of 2009 (Fig. 2a). The land consolidation project was executed from the au-
tumn of 2009 to the spring of 2010: Soil was cut and banked; old earthen ditches were buried; concrete
irrigation and drainage ditches were constructed; levees of paddy lots were readjusted (Fig. 2b). After
the project, there were 20 large lots (3,649m2 / lot) with concrete irrigation ditches with a width of
30cm and concrete drainage ditches with a width of 200-250cm (Fig. 2c). On the other hand, In OPA,
any construction work was not executed. There were still 36 small irregular-shaped lots (651m2 / lot)
with earthen ditches in OPA.
Before project(2008-2009) After project(2011-2013)
OPA IPA OPA IPA

200m
Survey lot
* **
* The lots were inaccessible because of cultivation abandonment.

**The lots were excluded from aggregate analysis because of no comparable data before project.
Fig. 1 Overview of study area
a b c
Fig. 2 Change of IPA (a: before the Project; b: during the project; c: after the project)
2.2 Target species

The brown frogs (Fig.3a), of which snout-vent-length is 3-7cm, live in the plains and hillsides.
They inhabit forests and grasslands in the non-breeding season, and migrate to the still water of paddy
fields, marshes, and small ponds for breeding (Maeda and Matsui 2002; Mori 2012). In the study area,
the breeding season was from late March to early May.
A mature female frog spawns only one jelly-like egg-mass (Fig.3b) during each breeding sea-
son (Maeda and Matsui 2002). Hence, the number of egg-masses is an indicator of that of mature fe-
males. Fluctuations in year-to-year egg-masses can be observed easily, and have been reported in
many nursery wetlands (e.g. Marunouchi et al. 2002; Osada 1978; Tomioka 2000; Biodiversity Center
of Japan 2009).
a b
Fig. 3 The Japanese brown frog (a) and its egg-mass (b)
2.3 Field survey

The survey was conducted 3-8 times in 2008-2013 (Table 1). 2008 and 2009 were before the
project; 2010 was during the project; 2011, 2012 and 2013 were after the project. Referring to Ha-
segawa (2003), we counted the egg-mass number, while walking along every paddy levee in IPA and
OPA. In 2010, however, the survey was conducted only in OPA, because IPA was shut out during
construction works. The number and location of egg-mass was recorded in a map. The map was used to
distinguish newly spawned egg-masses and old ones to avoid overlap at second survey downward in each
year.
Table 1 Outline of field survey
Survey
Project Year Times Survey date IPA OPA
before 2008 3 th th
Apr. 14 , 28 , May 9 th
○ ○
2009 5 th th
Mar. 28 , Apr. 10 , 20 , 27 , May 7 th th th
○ ○
during 2010 3 Mar. 12th, 30th, May 1st × ○
after 2011 6 rd th th
Mar. 3 , Apr. 5 , 14 , 26 , May 6 , 13 th th th
○ ○
2012 7 st th nd
Feb. 21 , Mar. 27 , Apr. 2 , 16 , 26 , May 2 , 11 th th nd th
○ ○
2013 8 th th th th
Mar. 18 , 28 , Apr. 5 , 17 , 25 , May 2 , 14 , 29 th nd th th
○ ○
3. Results and discussions
Fig. 4 shows the results of egg-mass survey in 2008-2013. Total of 319-1,435 egg-masses were
observed in IPA and OPA every year. In IPA, egg-masses were found at 62-78% of lots before the
project (2008-2009), while it was found at 15-50% of lots after the project (2011-2013).
Fig. 5 shows changes of egg-mass density (number / km) in IPA and OPA. Average size of the
egg-mass density in IPA was 66 before the project (2008-2009), while it was 7 after the project
(2011-2013). On the other hand, the size of the egg-mass density in OPA ranged 45-210, suggesting
the egg-mass density may naturally fluctuate even without the project. Hence, to correct for the influ-
ence of yearly fluctuation, the density ratio between IPA and OPA (= the density of IPA / that of OPA)
was calculated (Fig. 6). Average size of the density ratio was 0.5 before the project (2008-2009), while
it decreased to only 0.05 after the project (2011-2013). This suggests that decrease in egg-mass density
in IPA after the project was remarkable even though egg-mass density may naturally fluctuate as in
OPA.
Previous studies compared the numbers of brown frogs either among locations or be-
fore-and-after a project. Fujioka and Lane (1997) compared the numbers among locations, showing it
was low at paddy areas where a land consolidation project had been executed, as compared with those
where no project was executed. Hasegawa (1995) compared the numbers before and after a land con-
solidation project, showing it decreased after the project. Our study compared the numbers between
IPA and OPA before-and-after the project, suggesting that the frogs as well as the egg-masses would
diminish due to the project as Fujioka and Lane (1997) and Hasegawa (1995).
2008 2009 2010 Egg-mass

number
0
1～ 5
6～10
11～
No data
2011 2012 2013
200m
Fig. 4 Distribution of egg-masses in 2008-2013
200
Density (number /km)

OPA
100 Project
IPA
0
2008 2009 2010 2011 2012 2013
Year
Fig. 5 Changes of egg-mass density in IPA and OPA
0.60
Density ratio (IPA / OPA)
0.40 Project
0.20
0.00
2008 2009 2010 2011 2012 2013
Year
Fig. 6 Changes of density ratio

4. Conclusions
The egg-mass survey for 6 years suggests that the brown frogs in IPA would diminish due to the
project. The density ratio, however, seemed to slightly increase from 2011 to 2013 (Fig. 6). Hence, we
would continue the survey to verify whether the number of the brown frogs will vanish or recover
hereafter.
References
Biodiversity Center of Japan, 2009, Annual report of rural areas of Monitoring Sites 1000 Project,
Japan. (In Japanese)
Fujioka, M. and Lane, S.J., 1997, “The impact of changing irrigation practices in rice fields on frog
populations of the Kanto Plain, central Japan.” Ecological Research, 12, 101-108．
Hasegawa, M., 1995, “Nature of Yatsuda and brown frog.” In: Ohara, T. and Ohsawa, M. ed.
Bio-global science –Natural history of South Kanto-. Asakura Publishing Co. Ltd, 105-112. (In
Japanese)
Hasegawa, M., 2003, “Techniques in surveying organisms for agricultural engineers (8): field study of
amphibians.” Journal of the Agricultural Engineering Society, Japan, 71(5), 423-427. (In Japa-
nese)
Maeda, N. and Matsui, M., 2002, Frogs and toads of Japan, revised edition. Bun-Ichi Sogo Shuppan,
46-47.
Marunouchi J., Kusano T. and Ueda H., 2002, “Fluctuation in abundance and age structure of a
breeding population of the Japanese Brown Frog, Rana japonica Günther (Amphibia, Anura).”
Zoological Science, 19(3), 343–350.
Mori, A., 2012, “Analysis of rural ecosystem in Japan using stable isotope ratio.” Applied Photosyn-
thesis, 403-422.
Osada, K., 1978, “Population decline of Rana japonica in Kashiwa region. Natural history of frogs
(2).” Chiba Seibutsushi, 27(12), 102–107. (in Japanese)
Tomioka, K., 2000, “Long-term dynamics of breeding activity of two brown frogs Rana japonica and
R. ornativentris in the northern Kanto Plain, Japan.” Journal of the Natural History Museum and
Institute, Chiba, Special Issue, 3, 9–16. (in Japanese)
THE FUTURE
D-02
Applicability Study of Ecological Impact Assessment Using

AQUATOX Model in Paldang Reservoir, South Korea
Chun Gyeong Yoon, Han-pil Rhee *,†, Yeong-Kwon Son

Department of Environmental Science, Konkuk University, 1 1 Hwayang-dong, Gwangjin-gu,
Seoul, Korea
*ETwaters Inc., 93-3 Hwayang-dong, Gwangjin-gu, Seoul, Korea
**Rutgers School of Public Health, The State University of New Jersey, 683 Hoes Lane West,
Picataway, NJ 08854, U.S.
(†corresponding author : raphioss@konkuk.ac.kr; +82-2-455-3931)
Abstract
The purpose of this study was to evaluate the applicability of AQUATOX model, and to suggest the
methodology for practical use of ecological toxicity model in Korea. Paldang reservoir had been
selected as the study area, and BASINS/WinHSPF, watershed model, was set up for linked simulation
with AQUATOX in Paldang watershed. And then AQUATOX model was set up from the simulation
results of WinHSPF. AQUATOX was performed for analysis of ecological state and characteristics
including seasonal variation in Paldang reservoir. On the other hand, the ecological impact of
toxicants was predicted using scenarios according to inflow concentrations of alachlor, paraquat and
copper sulfate. In the case of alachlor, which have been classified as second degree of toxic material
to fish, the biomass of Minnow and Bass were declined considerably, when the concentration of inflow
was above 100 μg/L. In the case of paraquat, the biomass of minnow and bass were showed significant
declines for the inflow concentration of 10 μg/L. Copper sulfate showed a remarkable ecological
impact in spite of low inflow concentrations. AQUATOX model was evaluated as applicable to
waterbody in Korea, and can be used with existing various watershed and waterquality models
Keywords: AQUATOX, ecological toxicity model, ecological risk assessment, HSPF, water quality
model
1. Introduction
South Korea is a densely populated country with over 48 million people living in less than
100,000 km2. Until the late 1980s, rapid industrialization was the cause for serious damage to South
Korea’s natural environment. However, since then environmental conditions have been improving due
to the restoration programs. The ‘Special Act on Watershed Management for Four Major Rivers,’
enacted in 1998, includes a number of programs that aim to improve, maintain or restore the water
quality in national water systems (Rhee et al., 2012). These programs include discharge limits, permits
for point sources, funding for wastewater treatment facilities and a total daily maximum load (TMDL).
From 2010, over 90.1 % of the domestic wastewater generated nationally was collected and treated in
public sewers (MOE; Ministry of Environment, Republic of Korea, 2011).
Recently in Korea, the public interest for quality of life is increasing with improvement of the standard
of living and a rise in national income. And the social consciousness for water resource and water
quality, which can have a direct influence on life environment and vital activity, is changing quickly.
The river-planning oriented water resources management for securement of water quantity and
prevention of disasters had been a major concern until the late part of the last century. But now, water
quality management has become one of the most important social issues in Korea. After some famous
environmental accidents including water pollution accidents of Nakdong river by phenol (1991) and
organic solvent (1994), The public awareness about management of water quality and environment
was more diffused by active public relation programs including campaigns, education of Government
and NGO (non-government organization). And since late-2000s, nonpoint sources, aquatic life,
ecosystem and environmental toxicity were became key words of environmental policies.
The present environmental policies in Korea aim for “protection and restoration of sustainable, healthy
and sound environment,” and new systems and institution, including toxicological evaluation and risk
assessment of treated sewage, PRTR (pollutant release and transfer registers) and healthiness
evaluation of aquatic life, were established (MOE, 2008a).
Meanwhile, various modeling techniques have been applying as useful tools for development of water
quality management policies through deduction of proper control measures of pollutant in watershed,
the prior environmental review, prediction of future water quality and development of TMDL.
However, it is hard to support satisfactorily to recent target of policies by only current modeling
techniques. And actually, the most of them have been generally focused on physiochemical water
quality parameters, coliform, chl-a. Therefore, the application of new integrated modeling techniques,
which can predict the environmental fate of various pollutants and their effects on aquatic life, are
required.
AQUATOX model is a general ecological risk assessment model that represents the combined
environmental fate and effects of conventional pollutants, such as nutrients and sediment, and toxic
chemicals in aquatic ecosystems (US EPA, 2009). And this model can be applied through link with
BASINS (better assessment science integrating point and nonpoint sources), WinHSPF (The
hydrologic simulation program-Fortran) and SWAT (soil and water assessment tool), which have been
already used as national water quality modeling system in Korea. According to policy direction and
social requirement for water environment, AUQATOX seems like a very useful model in Korea.
However, the proper review about applicability is essential for introduction of new modeling system,
even though it is already verified model sufficiently in other countries. Furthermore, AQUATOX is
not yet commonly applied, and the form of required data is different from general ecological
monitoring data.
To address the aforementioned concerns, the applicability study of AQUATOX was performed, and
the methodology for practical uses was reviewed. For this reason, BASINS/WinHSPF model was
applied and simulated on Pladang watershed, and AQUATOX model was set up on Paldang reservoir
from the simulation results of HSPF through the BASINS AQUATOX Extension. The various
ecological data was collected and converted reasonably as biomass (AFDM; Ash-free dry mass) data
through literature review. Additionally, ecological impact by inflows of toxicants was predicted to
examine the utilizability of AQUATOX.
2.1 Study area
Paldang reservoir, which was selected as study area, is the major source of water supply of
Seoul metropolitan city and capital area, and Kyeong-an stream, North and South Han river flow into
this reservoir. Cheongpyeong dam is located on northeast direction, and Ipo weir is located on
southeast direction of reservoir (Figure 1). Paldang reservoir is riverine-type reservoir, and has
relatively small surface area (36.5 km2), storage capacity (244 million tons), average water depth (6.5
m) and short detention time (about 5.4 days) as compared with watershed area (23,800 km2) and
average inflow rate (44 million tons/day).
Paldang watershed includes parts of Gangwon-do, Gyeonggi-do and Chungcheongbuk-do provinces.
Total 33 unit watersheds for TMDL and 13 large dams, including Paldang dam are located in this
watershed. The average annual precipitation is about 1,270 mm, approximately 70% of which falls in
summer from July to September, with the remaining amount occurring from October to May. Due to
the Asian monsoon cycle, precipitation has large seasonal and spatial variation (Lee et al., 2010). Land
use types of the study area were based on MOE (Ministry of Environment, Republic of Korea) maps,
and total 23 land use types of original MOE land use map were reclassified into 7 categories; urban
area, agricultural land, forest, pasture, wetland, barren land and water (Table 1).
(a) Paldang watershed (b) Paldang reservoir
Figure 1. Location of paldang watershed and reservoir
Table 1. Land use distribution of the watershed
Urban Agricultural Forest Pasture Wetland Barren Water

Area (km2) 581.56 2996.58 19185.87 218.22 67.65 189.84 419.79
% 2.46% 12.67% 81.09% 0.92% 0.29% 0.80% 1.77%
2.2 Watershed modeling
The HSPF model (version 12.0) is a sophisticated continuous watershed model capable of
simulating hydrologic time series of runoff quantity-quality events (Bicknell et al., 2001). The
HSPF model can be applied to determine flows (hydrographs) and conventional pollutants
(pollute-graphs). Furthermore, the HSPF can be applied to the lumped parameter continuous
simulation model that has evolved out of Stanford Watershed Model, the US EPA agricultural runoff
management model, and non-point source model. HSPF can also be used as a distributed parameter
model, as it reproduces spatial variability by dividing a basin in hydrologically homogeneous land
segments and simulating runoff for each land segment independently (Lee et al., 2010). HSPF has
been used to simulate water flow and water quality for water resource management in Korea (Hwang,
2007; Im et al., 2003; Jeon et al., 2007; Jung et al., 2007). HSPF has been widely used for watershed
management to simulate various hydrologic conditions (Albek et al., 2004; Zarrillo and Ries, 2000),
transport of various nonpoint source pollutions including contaminated sediment (Donigian and Love,
2003; Hummel et al., 2003), land use management and flood control scenarios (Brun and Band, 2000;
Donigian et al., 1997).
In this study, HSPF model was selected for linked simulation with AQUATOX. The study area was
located among total 19 weather station of the Korea Meteorological Administration (Table 2). Hourly
weather data such as precipitation, air temperature, dew point temperature, wind speed, cloud cover,
solar radiation were obtained from those stations. And total 488 environmental Infra-facilities (74
terminal sewage disposal plants, 6 industrial wastewater treatment plants, 401 small scale sewage
systems, 5 excrements disposal facilities and 2 livestock wastewater treatment facilities) are located in
Paldang watershed (Figure 2-a).
Table 2. location of weather station in the study area (Paldang watershed)
Station code Station Name Latitude Longitude Elevation

95 Cheorwon 38°09' 127°18' 154.2
98 Dongducheon 37°54' 127°04' 112.5
100 Daegwallyeong 37°41' 128°46' 842.5
101 Chuncheon 37°54' 127°44' 76.8
108 Seoul 37°34' 126°58' 85.5
114 Wonju 37°20' 127°57' 149.8
119 Suwon 37°16' 126°59' 33.6
121 Yeongwol 37°11' 128°28' 239.8
127 Chungju 36°58' 127°53' 69.1
131 Cheongju 36°38' 127°27' 57.4
202 Yangpyeong 37°29' 127°30' 47
203 Icheon 37°16' 127°29' 77.8
211 Inje 38°03' 128°10' 198.6
212 Hongcheon 37°41' 127°53' 140.6
216 Taebaek 37°10' 128°59' 713.4
221 Jecheon 37°09' 128°12' 263.2
226 Boeun 36°29' 127°44' 174.1
272 Yeongju 36°52' 128°31' 210.2
273 Mungyeong 36°37' 128°09' 170.4
Each sub-basins was determined by natural drainage boundaries using a DEM (digital elevation
model) of MOE and “automatic delineation utility” in the BASINS, and was slightly modified to
calibrate HSPF model using monitoring data for TMDL of MOE. Based on the Korean Hydrologic
System, unit watershed for TMDL and Dams, Paldang watershed was divided into 50 sub-basins.
The calibration and validation for hydrologic simulation of HSPF model was performed at the points
of total 33 TMDL monitoring stations, and water quality was performed at the points of total 9 stations
(Figure 2-b). The period of calibration (2007 ~ 2008) and validation (2009 ~ 2010) were total 4 years,
and percent (%) difference (ASCE, 1993) were used to evaluate the model simulations, as follows:
(1)
Table 3 shows some general guidelines of %Diff. for calibration/validation tolerances or targets that
have been provided to model users in HSPF training workshops over the past 10 years (Donigian,
2000). The percent difference values of water quality are <15 ~ 35%, and the simulation results can be
judged as “Very good” to “Fair”.
Table 3. Criteria of percent difference (%Diff.) for evaluation of model simulations
Very Good Good Fair Poor
Water flow <10 10 ~ 15 15 ~ 25 -

Water quality <15 15 ~ 25 25 ~ 35 -
(a) Terminal sewage disposal plants (b) Station for model calibration and validation
Figure 2. Application of HSPF watershed model
2.3 Application of AQUATOX
Data from HSPF can be linked to the AQUATOX ecosystem model through the BASINS
AQUATOX Extension. AQUATOX accepts input for the water volume of the system, nutrient
time-series loadings, organic chemical time-series loadings, and physical characteristics of the system
(US EPA, 2005). AQUATOX model was set up on Paldang reservoir based on the simulation results
(geographic, meteorologic, hydrologic and water-quality data) of HSPF. Linkage function of BASINS
can support to generate and convert to appropriate data form from HSPF simulation results, and make
up for relatively vulnerable hydrologic simulation of AQUATOX.
Food web of AQUATOX has to consist of at least one aquatic life according to trophic levels, which is
categorized as “Plants (algae and macrophytes)”, “Invertebrates”, “Fish”. Plants category is
reclassified such as diatom, green algae, blue-green algae, other algae and macrophytes, and
invertebrates category is also reclassified as shredders, sed. feeders, susp. feeders, clams, grazers,
snails and pred. inverts. Fish category is reclassified as forage fish, bottom fish and game fish. Finally,
all of the ecological data about aquatic life must be reclassified to family or order levels of biological
classification, and input as biomass (AFDM; mg/L dry, g/m2 dry).
Ecological data (density; cells/L or ind./L, carbon contents; µg C/L) about plants and invertebrates of
Paldang reservoir were collected by literature review including report of NIER (2005). Rosemarin
(1982), Anderson (1995) and Stevenson et al. (1996) had considered that carbon contents (µg C/L) of
algae are about 36% of dry weight (mg/L dry) through their researches. And Bunting and Pretty
(2007) had also considered carbon contents of plants as 36% of dry weight. On the other hand, Pace
and Orcutt (1981) had considered that carbon contents of zooplankton are about 48% of dry weight
through their research. Therefore, biomass data as dry weight of plants and zooplankton categories
were converted from carbon contents using conversion factor (plants; 1/0.36, zooplankton; 1/0.48).
Phytoplankton, including Diatom, Green algae, Blue-green algae and Flagellate, and Myriophyllum
were selected as plants category. And zooplankton, including Copepod, Cladoceran and Rotifer, in the
invertebrates category were selected.
NIER (2006) had surveyed population and dry weight according to species of fishes in the Han river
system including Paldang reservoir. And major dominant species including Minnow, Bluegill, Carp
and Bass were selected for model simulation. Meanwhile, the population of zooplankton and
phytoplankton are affected by not only environmental conditions of waterbody and feeds but also
inflow or outflow, because of their planktonic characteristics on movement. On the other hand, fishes
have strong free movement capacity and relatively large body as compared with any other aquatic life,
and they can swim upstream. However, fishway of Paldang reservoir is blocked by Cheongpyeong
dam (direction of North han-river) and Ipo weir (direction of South han-river). Therefore, movement
or external loads of fishes were limited as below 3% (the portion of cross-sectional area of waterbody
at the reservoir boundary is less than this) of internal population in this study. Selected aquatic lives,
initial condition (biomass) and trophic interaction are summarized in Table 4. The information of
trophic interaction among aquatic lives was applied based on default values that are provided in
AQUATOX.
2.4 Senarios
Three toxicants including alachlor, paraquat and copper sulfate were selected to the review for
simulation of ecological impact in AQUATOX model. Until now, alachlor and paraquat have been
used widely and frequently as herbicide. It is well known that alachlor can cause adverse human health
effect including endocrine disrupting, injury of eyes, liver, kidney and spleen. MOE (2008b) reported
that alachlor had been detected several times at the ranged from 0.002 to 0.030 µg/L. paraquat
(paraquat dichloride) is well known as methyl viologen or Gramoxone. This herbicide has been very
commonly used for agricultural activity in Korea. It is known that LD50 (50% lethal dose) by
ingestion of paraquat is about 150 mg/kg (mouse), and LD50 by dermal contact is about 236 mg/kg
(rabbit). And It is well known through accidents of suicide by taking poison. Copper sulfates have
been used as the additives of most livestock feeds in Korea. Especially, gains rate of livestock is
improved when the copper sulfate is mixed at the rate of 125 ~ 250 mg/kg in feedstuff of pigs or
chicken. Therefore, copper sulfate can be discharged into water body with livestock manures.
Table 4. Initial condition (biomass) and trophic interaction among selected aquatic lives in AQUATOX
Initial Trophic interaction (preference, %)

condition
Cope Clado Roti Min Blue
(biomass) Carp Bass
pod ceran fer now gill
Refractory detritus of sediment -
Labile detritus of sediment - 22.2
Refractory detritus of particulate matter - 13.0 8.2
Labile detritus of particulate matter - 9.1 43.5 10.0 8.2 22.2
Diatom 0.553 mg/L dry 18.4
Greens 0.481 mg/L dry 18.4 5.6

Phytoplankton
Plants Blue-Greens 0.036 mg/L dry 54.5 18.4 5.6
Flagellate 0.380 mg/L dry 36.4 43.5 90.0
Macrophyte Myriophyllum 0.003 g/m2 day 22.2
Copepods 0.017 mg/L dry 14.3 100.0 11.1 1.5
Inverte
Susp Feeders Cladocerans 0.069 mg/L dry 14.3 11.1
brates
Rotifers 0.011 mg/L dry
Small forage Minnow 3.13 g/m2 dry 10.6
Large forage Bluegill 2.25 g/m2 dry 7.3

Fish
Large bottom Carp 0.58 g/m2 dry 14.7
Largemouth
Large game 0.30 g/m2 dry
Bass
Hypothetic scenarios for model application was made that inflow concentration of each toxicant is
increased exponentially as 5 levels (Control - 0, 1, 10, 100, 1000 ppb) from 0 to 1,000 ppb. This
scenario is impractical, but this can be advantageous to examine the extreme changes of biomass of
aquatic lives during relatively short model run times. Scenarios were applied to test whether ecological
effect and response of aquatic lives for toxicants is appeared reasonably, because this study focused on
evaluation of applicability and availability of AQUATOX. Acute toxicity data used in AQUATOX are
summarized in Table 5.
Table 5. Acute toxicity data of three toxicants used in the model
Exp. Time Alachlor Paraquat Copper sulfate
Aquatic life Endpoint
(hr) (µg/L) (µg/L) (µg/L)
Diatom 96 EC50 460 559 35
Green algae 96 EC50 460 0.55 5
Blue-green algae 96 EC50 4,600 1,120 85
Macrophyte 96 EC50 198 51 -
Copepod 48 LC50 35,003 11,680 -
Cladoceran 48 LC50 9,000 4,000 182
Rotifer 48 LC50 - - 76
Minnow 96 LC50 5,000 41,090 -
Bluegill 96 LC50 2,800 13,000 1,500
Carp 96 LC50 3,311 15,000 -
Bass 96 LC50 2,891 20,045 -
3. Results and discussion
3.1 Modeling results of HSPF

HSPF model was calibrated using 8-day interval observed data for Korean TMDL during 2007
~ 2008 and was validated during 2009 ~ 2010. LZSN, INFILT, KVARY, AGWRC, UZSN, INTFW
and IRC among many parameters related to hydrologic simulation are relatively sensitive. The
specificity of hydrologic simulation results became comparatively small, because each area of
sub-basins in this study was large-scale. Therefore, the adjustment of only three parameters (INFILT,
KVARY and AGWRC) was enough to gain acceptable level of calibration results. Each parameter are
adjusted finally as summarized in Table 6.
Table 6. The range of hydrologic parameters that were used for calibration of HSPF
Model Possible Initial Final

Parameter Description Unit
range range value valuse
0.0001 0.001
INFILT Index to infiltration capacity in./hr 0.16 0.10 ~ 0.27
~ 100.0 ~ 0.5
KVARY Variable groundwater recession 1/in. 0.0 ~ 5.0 0.0 ~ 5.0 0.0 0.00 ~ 0.8
0.001 0.85
AGWRC Base groundwater recession none 0.98 0.92 ~ 0.98
~ 0.999 ~ 0.999
According to model results of HSPF, the percent difference of hydrologic simulation was calculated as
0.01 ~ 12.55% in calibration, and as 0.0 ~ 12.06% in validation, respectively. And model efficiency
was evaluated as “Very good” for 28 sub-basins, and as “good” for 5 sub-basins.
Meanwhile, calibration and validation for water quality simulation, including water temperature,
dissolved oxygen (DO), biochemical oxygen demand (BOD), total nitrogen (T-N) and total
phosphorus (T-P), was performed for selected 9 points (Golji-A, Hangang-B, Hangang-D, Hangang-E,
Heukcheon-A, Bukhan-D, Jojong-A and Kyongan-B), And main parameters, which are used for
calibration, are summarized in Table 7. Generally, ELEV and LEVAP are used as important
parameters for calibration of water temperature. However, they were not used in this study, because it
was unnecessary.
Table 7. The range of parameters that were used for water quality calibration of HSPF
Model Initial Final

Parameter Definition Unit
range value value
1.0E-30 0.004
KBOD20 Unit BOD decay rate at 20℃ 1/hr 0.004
~ none ~ 0.067
0.011
KODSET BOD settling rate ft/hr 0 ~ none 0.027
~ 0.027
Empirical constant in the equation used to 1.0E-30 0.05
REAK 1/hr 0.2
calculate the reaeration coefficient ~ none ~ 0.2
0.0001 0.001
BRBOD Base release rate of BOD materials mg/m2 0.001
~ none ~ 150.0
Conversion from milligrams biomass to
CVBO mg/mg 1.0 ~ 5.0 1.63 1.63
milligrams oxygen
Conversion from biomass expressed as
CVBPC moles/mol 50 ~ 200 106 106
phosphorus to carbon
Conversion from biomass expressed as
CVBPN moles/mol 10 ~ 50 16 16
phosphorus to nitrogen
BPCNTC Percentage of biomass which is carbon none 10 ~ 100 49 49
The benthal release rates of ammonia 0.0
BRNIT mg/m2 0.0 0.0 ~ 95
under aerobic and anaerobic condition ~ none
0.001
KTAM20 Nitrification rates of ammonia at 20℃ 1/hr 0.045 0.045
~ none
0.001
KNO220 Nitrification rates of nitrate at 20℃ 1/hr 0.008 0.008
~ none
0.001
KNO320 Denitrification rates of nitrate at 20℃ 1/hr 0.008 0.008
~ none
0.001
MALGR Algal growth rate 1/hr 0.0085 0.085
~ none
Table 8. The evaluation of model efficiency after calibration and validation of HSPF
Calibration Validation
%differ. Efficiency %differ. Efficiency
Flow 0.01 ~ 12.55 Very good ~ Good 0.0 ~ 12.06 Very good ~ Good
Water Temperature 1.46 ~ 9.41 Very good 1.15 ~ 8.54 Very good
DO 0.33 ~ 7.74 Very good 0.27 ~ 11.01 Very good
BOD 0.87 ~ 15.43 Very good ~ Good 2.3 ~ 18.69 Very good ~ Good
T-N 2.44 ~ 8.15 Very good 1.25 ~ 7.97 Very good
T-P 5.41 ~ 17.39 Very good ~ Good 10.7 ~ 25.33 Very good ~ Fair
According to model results, percent differences for water quality simulation were calculated and
evaluated as summarized in Table 8. And Figure 3 shows example sub-basins (Kyeongan-B) among
calibration/validation results for hydrologic and water quality simulation of HSPF.
(a) flow (b) Water Temperature
(c) DO (d) BOD
(e) T-N (f) T-P

Figure 3. Example of calibration/validation result (Kyeongan-B) of HSPF
3.2 Modeling results of AQUATOX

Figure 4 shows the comparison between observed data at the outlet of Paldang reservoir and
simulated results on flow and general water quality of AQUATOX. According to the evaluation of
model efficiency using percent difference, the results of flow and general water quality were fairly
congruent (Table 9). It is noteworthy that the validation process for two independent data sets was
successful.
Table 9. The evaluation of model efficiency of AQUATOX
Calibration Validation
%differ. Efficiency %differ. Efficiency
Flow 6.84 Very Good 2.28 Very Good
Water Temperature 0.16 Very Good 4.02 Very Good
DO 5.36 Very Good 1.01 Very Good
BOD 12.66 Very Good 13.63 Very Good
T-N 5.77 Very Good 13.58 Very Good
T-P 6.77 Very Good 7.53 Very Good
SS 29.74 Fair 17.02 Good
Chl-a 11.21 Very Good 9.12 Very Good
(a) flow (b) Water Temperature
(c) DO (d) BOD
(e) T-N (f) T-P
(g) SS (f) Chl-a

Figure 4. Flow and general water quality simulation of AQUATOX
Before the application of scenario for toxicant inflow, control condition of AQUATOX had been
simulated during 5 years (2005 ~ 2010), and initial 3 years (2005 ~ 2007) were considered as period
for model stabilization. Biomass of algae and zooplankton was increased from early spring to autumn.
Planktonic Biomass seems to be affected by climate, nutrient inflows and hydrologic variation.
However, food web is also very important factor to biomass variation, especially biomass of
zooplankton, including copepods, cladocerans and rotifers, which are feeding on flagellate commonly,
seem to be related on flagellate, not diatom or green algae (Figure 5). The biomass of minnow, which
prefer green algae and diatoms, was increased until autumn with its’ feeds. Carp, which is preferred by
bass, showed decrease of biomass with increase of bass. And variation of bass can be also related with
minnow. Modeling that food web is considered is very complex, but it can provide comprehensive
information.
(a) Phyto., Diatoms (b) Phyto., Green algae
(c) Phyto., Blue-green algae (d) Flagellate
(e) Copepods (f) Cladocerans
(g) Rotifers (f) Minnow
(i) Carp (j) Largemouth bass

Figure 5. Flow and general water quality simulation of AQUATOX
3.3 Application of Scenarios
AQUATOX was simulated to examine the reflection of ecological impact in this model,
especially variation of biomass according to inflow concentration of toxicants. In the case of alachlor,
planktonic biomass did not show significant variation. Algae and zooplankton were come into the
reservoir continuously from upstream during the period of simulation. Therefore, it is not easily
outstanding in the total budget, even though a portion of planktonic biomass was decreased.
Furthermore, EC50 of plants category is set up as from 460 to 4,600 µg/L (LC50; 14,290 ~ 142,900
µg/L), more highly than inflow concentration of scenarios. However, biomass of fishes, including
minnow, carp and bass, showed significant variation (Figure 6).
(a) Phyto., Green algae (b) Minnow
(c) Carp (d) Largemouth bass

Figure 6. The variation of biomass according to inflow concentration of alachlor
(a) Minnow (b) Bluegill
(c) Carp (d) Largemouth bass

Figure 7. The variation of biomass according to inflow concentration of paraquat
In the case of paraquat, planktonic biomass showed similar results with alachlor, although LC50
and EC50 of paraquat are lower than alachlor (Figure 7). It seems that was caused by difference of
elimination rate constant. Algae and zooplankton have relatively higher elimination rate constant
(11,252 ~ 391,085) on paraquat than alachlor. Biomass of minnow and bass decreased definitely, but
bluegill and carp even increased. Bluegill has the most high elimination rate constant on alachlor and
paraquat (about 18 times of bass). Furthermore, it has relatively low lipid fraction.
Before simulation of copper sulfate, some aquatic lives (copepods, minnow, carp and bass) were
excluded, because of lack of toxicity data. According to simulation results, planktonic biomass
decreased as compared with alachlor and paraquat. Especially, biomass of zooplankton decreased
definitely, although decrease of predator and continuous inflow from upstream (Figure 8). Bluegill
also decreased at the condition above 10µg/L, and it seems to be hard to survive at the condition above
100µg/L. Algae are generally affected earlier than zooplankton or fishes, when copper sulfate came
into waterbody. This simulation results were caused by hydraulic characteristics (riverine-type
reservoir) of Paldang reservoir and biological data set up in the model.
(a) Phyto., Diatoms (b) Rotifers
(c) Cladocerans (d) Bluegill

Figure 8. The variation of biomass according to inflow concentration of copper sulfate
4. Conclusions
This study was initiated to evaluate applicability of AQUATOX, and to suggest the
methodology for practical use of ecological toxicity model in Korea. BASINS/WinHSPF watershed
model was applied to Paldang watershed. And AQUATOX model was set up by the liked simulation
data from HSPF. The various ecological data was collected and converted as available biomass data in
the model through literature review. AQUATOX model was performed for prediction of ecological
states and characteristics including seasonal variation associated with trophic levels in Paldang
reservoir. And the ecological impact by inflow of toxicants was also predicted using hypothetic
scenarios.
The conclusions, which obtained through this process, are as follows;
(1) AQUATOX model is applicable to waterbody in Korea and can be used with the linkage of other
established watershed and water quality models for Korean TMDL.
(2) However, the continuous accumulation of time series ecological data as the biomass (dry weight)
form and monitoring data of various toxicants in waterbody are essential to give a guarantee for the
reliability of prediction.
① Continuous ecological data series; biomass data (mg/L dry, g/m2 dry) according to biological
classification of phytoplankton, periphyton, macrophytes and invertebrates including zooplankton at
the important point of waterbody
③ Continuous biomass data (g/m2 dry) for the main dominant species of fishes
④ Continuous monitoring data of toxicants (µg/L) from waterbody and chemical discharge facilities
including PRTR (pollutant release and transfer registers) data
(3) And if the weakness for hydraulics is supplemented with other suitable model such as WASP and
EFDC, AQUATOX will become more available for sustainable water resource management through
various prediction and analysis.
AQUATOX model has relatively high compatibility with existing watershed model for TMDL such as
HSPF and SWAT. To reflect the recent state of Korea, which was formed a social consensus about the
importance of healthy ecosystem and toxicologically safe water resources, AQUATOX is worth
applying. It is expected that AQUATOX is usable as scientific tool for safe and sustainable water
resource management, if this model is properly used to predict ecological impact and risk.
Acknowledgements
This paper was supported by Konkuk University in 2013
References
Albek, M., Ogutveren, U. B. and Albek , E., 2004, Hydrological modeling of Seydi Suyu watershed
(Turkey) with HSPF. Journal of Hydrology, 285, 260–271.
Anderson L. A., 1995, On the hydrogen and oxygen content of marin phytoplankton. Deep-Sea
Research, 42(9), 1675-1680.
ASCE Task Committee on Definition of Criteria for Evaluation of Watershed Models, 1993, Criteria
for evaluation of watershed models. Journal of Irrigation and Drainage Engineering, 119 (3),
429–442.
Bicknell, B. R., Imhoff , J. C., Kittle, J. L., Jobes, T. H. and Donigian Jr., A.S., 2001, HSPF, Version
12, User’s Manual. US Environmental Protection Agency, Athens, GA, 2001, 120–128.
Brun, S. E. and Band, L. E., 2000, Simulating runoff behavior in an urbanizing watershed, Computers,
Environment and Urban System, 24 (1), 5–22.
Bunting, S. W. and Pretty, J., 2007, Aquaculture Development and Global Carbon Budgets :
Emissions, Sequestration and Management Options. Centre for Environment and Society,
Department of Biological Sciences, University of Essex, Colchester, UK.
Donigian Jr., A. S., 2000, HSPF Training Workshop Handbook and CD, Lecture #19, Calibration and
Verification Issues. Slide #L19-22, EPA Headquarters, Washington Information Center,
Presented and prepared for U.S. EPA, Office of Water, Offi ce of Science and Technology,
Washington, DC
Donigian Jr., A. S., Chinnaswamy, R. V. and Jobes, T. H., 1997, Conceptual Design of Multipurpose
Detention Facilities for Flood Protection and Nonpoint Source Pollution Control. Aqua-terra
Consultants, Mountain View, CA, 151
Donigian Jr., A. S. and Love, J. T., 2003, Sediment calibration procedures and guidelines for
watershed modeling. WEF Specialty Conference Proceedings on CD-ROM, Chicago, IL.
Hummel, P. R. Kittle, J. L., Duda, P. B. and Patwardhan, A., 2003, Calibration of a Watershed Model
for Metropolitan Atlanta. WEF Specialty Conference Proceedings on CD-ROM, Chicago, IL.
Hwang, H. S., 2007, Applicability study of BASINS/WinHSPF on TMDL in Korea. Ph.D. dissertation,
Department of Rural Engineering, Konkuk University, Seoul, Korea.
Im, S. J., Brannan, K. M., Mostaghimi, S. and Cho, J., 2003, A comparison of SWAT and HSPF
models for simulating hydrologic and water quality responses from an urbanizing watershed.
2003 ASABE Annual International Meeting Sponsored by ASAE Riviera Hotel and Convention
Center, Las Vegas, NV.
Jeon, J. H., Yoon, C.G., Donigian Jr., A. S. and Jung, K. W., 2007, Development of the HSPF-paddy
model to estimate watershed pollutant loads in paddy farming regions. Agricultural Water
Management, 90 (2007), 75–86.
Jung, K. W., Yoon, C. G., Jang, J. H. and Kim, H. C., 2007, Quantitative estimation of pollution
loading from Hwasung watershed using BASINS/HSPF. Journal of the Korean Society of
Agricultural Engineers, 49(2), 61–74.
Lee, S. B. Lee, Yoon, C. G., Jung, K. W., Jang, J. H. and Rhew, D. H., 2010, Watershed-scale
modeling to estimate delivery ratio of pollutant loads to support TMDL application in Korea.
Desalination and Water Treatment, 19 (2010), 64-73.
MOE, Ministry of Environment, Republic of Korea., 2011, Statistics of Sewerage 2010 [Internet].
Gwacheon: Ministry of Environment; c2010 [cited 2012 Mar 9]. Available
from: http://webbook.me.go.kr/DLi-File/086/5508730.pdf
MOE, Ministry of Environment, Republic of Korea., 2008a, Research on the actual condition of
provisional managed toxicants in the Hanriver basin, and development of prediction system.
MOE, Ministry of Environment, Republic of Korea., 2008b, Investigation of toxicants for expansion
of management target, and development of control measure.
NIER, National Institute of Environmental Research., 2005, Trophic dynamics and energy flow of
plankton food web in Lake Paldang. Available from:
http://www.hanriver.re.kr/App_JSP_Page/tb_resc_info.jsp?s_resc_grp_cd=1
NIER, National Institute of Environmental Research., 2006, Survey on Environment and Ecosystem of
Lakes in the Han river.
Pace, M. L. and D. Orcutt, Jr., 1981, The relative importance of protozoans, rotifers, and crustaceans
in a freshwater zooplankton community, Limnol. Oceanogr., 26, 822-830.
Rhee, H. P., Yoon, C. G., Lee, S. J., Choi, J. H. and Son Y. K., 2012, Analysis of nonpoint source
pollution runoff from urban land uses in South Korea. Environmental Engineering Research,
17(1), 1-10.
Rosemarin, A. S., 1982, “Phosphorus nutrition of two potentially competing filamentous algae,
Cladophora glomerata (L.) Kutz. and Stigeoclonium tenue (Agardh) Kutz. from Lake Ontario”, J.
Great Lakes Res., 8, 66-72.
Stevenson, R. J., Bothwell, M. L. and Lowe, R. L., 1996, Algal Ecology, Freshwater Bentic
Ecosystmes. Academic Press, An imprint Elsevir, CA, USA.
US EPA, 2005, AQUATOX (RELEASE 2.1); Modeling environmental fate and ecological effects in
aquatic ecosystems (Volume 3: User’s manual for the BASINS extension to AQUATOX release
2.1). EPA-823-B-05-001, Office of Water, U.S. Environmental Protection Agency, Washington
D.C.
US EPA, 2009, AQUATOX (RELEASE 3); Modeling environmental fate and ecological effects in
aquatic ecosystems (Volume 2: Technical documentation). EPA-823-R-09-004, Office of Water,
U.S. Environmental Protection Agency, Washington D.C.
Zarrillo, P. J. and Ries, K. G., 2000, A Precipitation-Runoff Model for Analysis of the Effects of Water
Withdrawals on Streamflow, Ipswich River Basin. Massachusette USGS Waters Resources
Investigations Report 00-4029, 99.
THE FUTURE
D-05
Feasibility Analysis of Nitrogen Balance in Paddy Fields toward

New Irrigation Service for Rice Quality
Tasuku Kato* and Toshiaki Iida**

*Tokyo University of Agriculture and Technology
** The University of Tokyo
Abstract
In irrigation for paddy fields, rice quality is influenced by water quality, because Japanese market
prefers lower nitrogen contents ratio in rice crop for the quality of taste. However, ordinary paddy
fields are located where population density is high around large city area for easy to access the large
market. Though irrigation system is managed by water user’s group that is composed by paddy
farmers, distribution of irrigation water is mainly focused on the equity in aspect of water volume. For
further new irrigation schema, distribution should be considered regarding with water quality.
In this paper, to develop purified water service, a feasibility analysis was conducted in degraded
irrigation water quality area, toward increasing demand oriented irrigation service. As study case,
Imbanuma basin was chosen because of large paddy area adjacent to city area, and water quality has
ever been degraded in this area. The target water quality index is total nitrogen concentration because
TN concentration influenced to protein contents ratio and taste quality. The nitrogen purified
mechanism in paddy fields is already understood as denitrification by microbiological activity. The
drainage water from paddy fields is assumed to collect and reuse in paddy fields, however, availability
of amount and quality is not analyzed quantitatively. As a feasibility test, nitrogen balance would be
analyzed. And as conclusion, it will be evaluated a coverage area for purified water irrigation.
Keywords: Demand oriented Irrigation; Water users group; Nitrogen balance; Water quality
1. Introduction
Nowadays, in aspect of international trade and water security, it is asked for the efficiency to
water utilize of irrigation in paddy fields. To implement water management in paddy field irrigation,
several approaches would be required, i.e. farming technology, engineering, institutional and regional
planning methods regarding water resources management should be arranged and designed on a new
paradigm. Until now in Japan, paddy field irrigation was conducted with supply oriented service, that
is, irrigation schedule is mostly governed by the water users institution, and farmers follow their
distribution schedule. In aspect of increasing service, water distribution should be more elastic and
sophisticated system that is highly reliable and functional. For example, information and
communication technology would be useful to manage agricultural production or food quality control
technology. In water users institution, those management way is considered in future.
In this paper, to develop purified water service, a feasibility analysis was conducted in degraded
irrigation water quality area, toward increasing demand oriented irrigation service. The target water
quality index is total nitrogen concentration because TN concentration influenced to protein contents
ratio and taste quality. The nitrogen purified mechanism in paddy fields is already understood as
denitrification by microbiological activity. The drainage water from paddy fields is assumed to collect
and reuse in paddy fields, however, availability of amount and quality is not analyzed quantitatively.
As a feasibility test, nitrogen balance would be analyzed.
2.1 Study Area and Monitoring
The study area is located in paddy fields area in

Imbanuma lake basin, Chiba, Japan. This lake is eutrophic
lake of municipal wastewater and agricultural drainage. The
five monitoring station were installed in the sub watershed,
Kashima river watershed, including paddy fields area (Fig. 1).
In monitoring station, hydrological data and water quality
parameters are observed.(Fig. 2 and 3)
2.2 Water Balance
First, continuous monitoring of hourly water level and

daily water quality was conducted, then, water balance and
nitrogen balance were estimated. Finally, availability of
purified water for paddy irrigation was evaluated based on the
water and nitrogen balance results. The boundary of the
estimation was settled separately river and sub watershed,
respectively.For river water balance equation was below.
Q10 + Q28 + Q13 － Q25– Ir + Dc – Rc = 0 (1)
where Q10,13,28 (summarized Qin) was inflow for each Fig. 1 Study area
monitoring point, that was estimated rating curve of water level and observed discharge volume. Q25
was outflow (Qout) volume. Ir was irrigation amount pumped up from river water. Dc was discharge
from watershed to river channel, and Rc was recharge volume from river to watershed. Dc was
contained both paddy field and background area discharge.
In this equation, Dc and Rc was not clearly defined, then, Dc was assumed on the storage
function method, and water balance in sub watershed scale was settled below:
P ＋ Ir – ET + Rc – Dc = dSw (2)
Dc=BG+D=K Swα =(k+k’)Swα (3)
where P was rainfall, ET was evapotranspiration, dSw was storage change in the sub watershed. BG
was discharge from background area and D was discharge from paddy fields. And K,k,k’,α were
parameters for storage function.
Rainfal Discharge mg/L mm/d
No.25(N)
日降雨量比流量
60 0
排水路比流量
12 60
5
50
10 10 50
40 15
8 40
20
30
25
6 30
20 30 4 20
35
10 2 10
40
0 45 0 0
7/26 8/26 9/26 10/26 11/26 12/26
7/18 8/18 9/18 10/18 11/18 12/18
Fig.2 Rainfall and discharge（2012/7/26～12/31）
Fig.3 Water quality (T-N) in discharge and drainage channel
2.3 Nitrogen Balance
Regarding nitrogen balance, the rating curve was calculated on water quality observed data and
discharge volume.
L=Cobs·Qobs (4)
where Cobs is observation concentration of total nitrogen (T-N), and Qobs is observed discharge.
Based on L, the logarithmic rating curve between Lout and Qout was developed by least squares
method.
Lout=αQoutβ (5)
where α,β are parameters of regression model.
3. Results and Discussion Table 1 Water Balance (up)，Nitrogen

Balance (middle)，average nitrogen
Qin Qout Ir D(d Rc) BG Rain ET
3.1 Water and Nitrogen balance 灌漑
Irrigation 5.70 7.46 0.23 0.33 0.98 2.64 0.93 3.83
期
非灌
Non-Irrig
The results showed in Table1. Based ation漑期
22.17 28.00 0.01 2.06 8.06 11.84 13.94 5.03
on these water and nitrogen balance,
virtually averaged water quality was Lin Lout L(Ir) L(Rc) L(D) L(BG) L(dSw) L(P)
estimated in Table1. The result of nitrogen
water quality was 1.3 mg/l that was lower Irrigation
灌漑期 19.16 16.20 0.77 3.31 0.64 0.25 3.41 0.23
than 3.8 mg/l in averaged river water quality. 非灌漑
Non-Irrig
79.66 71.58 0.05 28.95 0.00 17.45 15.01 3.46
This is the influence of denitrification in ation 期
paddy fields area. Also, discharge volume

Cin Cout C(Ir) C(Rc) C(P) C(D) C(BG)
was 0.33×106m3 that was assumed to be
Irrigation
灌漑期 3.36 2.17 3.36 3.36 0.25 1.95 0.10
available to around 40 ha for 20 mm/day
consumption in paddy fields. This area was 非灌漑
Non-Irrig
3.59 2.56 3.59 3.59 0.25 0.00 1.47
ation 期
1/10 of sub watershed paddy fields area
(around 400 ha).
3.2 Nitrogen purification effect to residential people
For development new irrigation service, nitrogen water quality change was observed and, in
irrigation period, nitrogen purification was confirmed. This effect is better to the residential people for
increasing amenity that is expected further environmental education or involvement for watershed
conservation planning process. Especially, lake eutrophication is problem in this study area and
prevention of eutrophication is expected to improve drinking water quality and urban environment.
This effect would be covered watershed scale wide, and watershed model, i.e SWAT model, is
helpful to analyze of water management.
3.3 Nitrogen purification effect to farmers
Rice taste quality is influenced by nitrogen contents ratio in rice. And nitrogen contents ratio
was majorly dominated by nitrogen supply amount after booting or during heading period. Also,
Nitrogen was supplied by not only fertilizer but also irrigation water quality. In this study area,
nitrogen water quality in irrigation water is relatively high and rice quality is not so high evaluated in
market.
If good water quality is available for paddy fields in irrigation period, rice quality would be
better. And paddy drainage water quality is better than irrigation water because of denitrification. Then
we would like to propose new water management improve good water quality irrigation water by
reuse system of drainage water. In those system, real time nitrogen control system is helpful by ICT
technology. Also to install the system, as amount of good water quality is limited, farmers are required
to communicate for reuse of drainage water. Agent based model or system dynamics approach would
be helpful for further analysis.
4. Conclusion
The possibility of a novel service for farmers were presented, in which the farmers can select
the timing of irrigation intake based on the information on water quality at irrigation canals.
Discussion should be conducted about water distribution framework in the demand oriented water use
scheme that includes reuse irrigation of purified water for rice quality. For example, the purified water
should not be distributed equally and freely. The purified water would be distributed to famers that are
willing to pay additional cost for water supply. It is one method to encourage the change to demand
oriented water supply system.
In future, the discussion should be conducted about water distribution framework in the demand
oriented water use scheme that was included reuse irrigation by purified water for rice quality. For
example, the purified water should not be distributed equally and freely, this purified water would be
distributed to famers that were available to pay additional cost for water supply. It is one method to
introduce to change to demand oriented water supply system.
Acknowledgments
This research was supported “Development of demand oriented water supply system and
evaluation of irrigation service” through Research Institution of Science and Technology for Society
(RISTEX) funded by Japan Science Technology agency (JST).
THE FUTURE
D-07
Relational Analysis between Yield and Planting Condition of

Rainy Season Rice in Low Productive Fields: a Case Study in Lao
PDR
Hiroshi Ikeura*, Phetyasone Xaypanya**, Sengthong Phongchanmixay***,

Somphone Inkhamseng**, Somnuck Soubat***, Salermphon Phonangeone**,
Soulintha Chanthabuly**
*Rural Development Planning Division, Japan International Research Center for Agricultural
Science 1-1 Owashi, Tsukuba, Ibaraki, 305-8686, Japan, e-mail:hiroikeura@affrc.go.jp, Tel:
+81-29-838-6687, Fax: +81-29-838-6693
** Water Resources Engineering Department, Faculty of Engineering National University of
Laos Tadthong Village, Sikhotaboong District, Vientiane Capital, Lao PDR,
e-mail: pxaypanya@yahoo.com
***Agricultural Research Center Naphok, National Agriculture and Forestry Research
Institute, Ministry of Agriculture and Forestry, Laos Thadindeng Road, Thadockham Village,
Xaythany District, Vientiane Capital, Lao PDR, e-mail: phongchanmixays@gmail.com
ABSTRACT
Rice productivity is different for each field, and low-productivity fields are existed in each village.
The differences in rice productivity are caused by differences in water use, cultivation method, plant-
ing environment, and other factors. In order to improve rice productivity in low-productivity fields,
the factors that influence yield reduction must be clarified. In this study, analysis of relationships
among farming activities, convenience of water use, and yield of rainy-season rice was conducted in
the northern part of Lao People’s Domestic Republic (Laos). A field survey was conducted in a vil-
lage in Feuang District, Vientiane Province, during the rainy season of 2012. The timing of farming
activities was recorded weekly in each lowland rice field from the end of May to the beginning of Au-
gust, when transplanting was completed in most of the fields. Yield surveys were also conducted in
29 fields located in two small river areas. Transplanting was primarily carried out in the mid-July,
one or two weeks later than in 2011. Rice yield varied from 2.4 t/ha to 6.8 t/ha; the low-productivity
fields were located in both downstream and upstream areas. However, no significant relationship
was observed between rice yield and location of the fields or water-use characteristics. Although the
difference was not significant, a decreasing tendency of yield was noted for late transplanting times.
Keywords: low-productivity field, difference of rice yield, time of transplanting, water-use situation
1. Introduction
Rice production in Laos has been increasing year after year and self-sufficiency in rice was
achieved at the national level in 2006 (WFP, 2007). The government of Laos is targeting further
progress in rice production aimed at exporting the rice abroad. According to Ministry of Agriculture
and Forestry statistics, lowland rice production in the rainy season accounts for about 77% of total rice
yield and plays an important role in sustaining food self-sufficiency in Laos. However, differences in
the level of self-sufficiency in rice exist among provinces (WFP, 2007). In addition to the differ-
ences in rice productivity at the national level, local differences also exist and are caused by water use,
planting environment, and other factors. In Laos, 22% of the population are still undernourished (FAO,
2011).
The difference in rice productivity is caused by differences in water use, cultivation methods,
and planting environment. In order to improve rice productivity in low-productivity fields in each
village, the factors that influence yield reduction must be clarified.
Ikeura et al. (2012) confirmed that farmers required a large amount of water for transplanting,
primarily in July, through interviews in 41 villages in Laos. Ikeura et al. (2012) also conducted a sur-
vey of planting schedule for lowland rice in the rainy seasons of 2010 and 2011 in a village in Vienti-
ane Province. The results of this interview survey indicated that most farmers transplanted rice seed-
lings in early July in both years. However, although competition for water and labour at the time of
transplanting were suggested from the results, the relationship between differences in rice yield in
each field and factors influencing productivity still have not been clarified.
In order to clarify these influential factors, in particular water condition and farming activities in
each field, an analysis of relationships between farming activities, convenience of water use, and yield
of rainy-season rice was conducted.
2.1 Survey site

Vientiane Province
The project site was N Village, Feuang
District, Vientiane Province, Laos. The village
is located 88 km from Vientiane municipality
towards the northwest. Fig. 1 shows the loca-
tion of the village. This village is located in a
lowland area between mountains and represents
a typical rice-planting area in northern Laos. Nam Ngum Dam
The population and number of households
in the Village in 2011 were 683 and 135, respec-
tively. There were 81 ha of rice fields in the
lowland area (lowland fields) and 12 ha of nar-
row paddy fields in the mountainous areas.
Fig. 2 shows the map of lowland fields in
N Village. All the fields were cultivated for
planting rice only in the rainy season. Half of
the lowland fields are cultivated by farmers that
live in the surrounding villages. The lowland Feauang
fields are divided into 5 areas based on water
flow from 5 small rivers. Nam Ngum
The beneficial area of River A (hereinafter N Village
Dam
referred to as ‘Area A’) was developed in the
5km
1970s and has the longest history in the lowland
fields of N Village. The other fields were de-
veloped year by year after the 1970s. Area A
Fig. 1 Location of survey site
and the beneficial area of River B (hereinafter referred to as ‘Area B’) are located near the village,
most of the fields were cultivated by farmers belonging to N Village. Areas A and B do not have water
allocation systems and irrigation water is supplied by plot-to-plot from upstream fields to downstream
fields.
2.1 Field survey

A field survey was conducted in Areas A and B from 28 May to 4 August 2013. There were
53 fields in the 2 areas, 35 of which were owned by farmers in N Village (Fig. 3). The farming ac-
tivities, such as establishment of nursery, ploughing, paddling, transplanting and direct seeding were
recorded weekly for each field. The water supply situation for each field, such as direct irrigation
from the canal or plot-to-plot through other owners' fields, was also confirmed.
River B
River A
River C
River D
River E
Canal Reservoir, Pond

Expected beneficial area of each water source River A River B River C River D River E
Fig. 2 Lowland rice field in N Village and survey areas
Yield survey was conducted
in 31 fields that were cultivated by
farmers from N Village. Three
quadrats (1 m × 1 m) were installed
in each field and samples were
collected from October to Novem-
ber when rice was mature. In
Field 46, 6 quadrats were installed
in 2 blocks in which 2 different rice
varieties were cultivated. Two
owners harvested rice from within
the quadrats prior to the sampling.
Ultimately, 90 samples were col-
lected from 29 fields. Paddy
weight of the samples was meas-
ured after air drying, threshing, and
winnowing. Moisture content of
the rice was also measured.
Fig. 3 Numbering of survey fields
Owned by farmers of the other village Rectangular weirs with
automatic water-level gauges (On-
The Fields numbered with red character are sampling fields
set HOBO U-20) were installed at
 The fields which can be taken water from canal the main outlets of the reservoirs of
River A and River B on 29 June 2012 to measure water discharge. Water level was measured every
5 minutes and was converted to discharge by using the Francis formula, shown below as Formula (1):
(1)
where, Q represents water discharge (m3/min); b is width of the weir outlet (m); and h is overflow
depth of the weir (m).
Rain gauges (Onset HOBO RG-3M) were installed in the village in February 2012. Climate
data observed at Feuang Station were obtained from the Department of Meteorology and Hydrology,
Laos (hereinafter referred to as ‘DoMH’) to compare the precipitation between 2012 and past years.
Table 1 Monthly precipitation in Feuang District
Avg.
2011 2012
(2007-2012)
3.1 Precipitation in 2012 Jan 1.3 5.3 19.9
Table 1 shows monthly precipitation in Feb 6.1 10.4 6.4
Feuang District in 2011, 2012, and the average Mar 152.7 21.2 54.0
of 6 years (from 2007 to 2012). Average an- Apr 171.3 102.3 119.0
nual precipitation was 2,380mm. Precipitation May 476.3 326.2 337.0
exceeded 2800 mm in 2011, a year in which Jun 568.4 208.8 361.2
there was heavy rain and extensive flooding in Jul 285.9 418.7 355.4
the Indo-China area. In 2012, there was less Aug 435.5 511.5 475.6
rainfall than both the average and the 2011 Sep 559.3 348.8 406.8
rainfall. Cumulative precipitation before Oct 181.7 77.3 194.5
transplanting from March to June in 2012 was Nov 39.5 92.3 41.0
lower, and it was half of that in 2011. Less Dec 3.4 5.7 7.0
precipitation occurred in 2012, resulting in Total 2,881.4 2,128.5 2,377.8
Based on the data observed at Feuang station by
poorer conditions for nursery establishment and
DoMH.
transplanting compared to 2011 and the average.
There was heavy rain in July and August, consistent with a typical year.
3.2 Time of farming activities for rice planting
Fig. 3 shows the times of nursery establishment, ploughing, paddling, transplanting, and direct
seeding were conducted in each field. When the activities were continued for several weeks,
Fig. 3 Time of farming activities and daily precipitation from 28 May to 4 August
they were counted two or more times.
The nurseries established outside of Areas A ~ 07 July
and B were not counted. Of 53 survey fields, ~ 14 July
transplanting was not completed in 5 fields by ~ 21 July
4 August and in 1 field, rice was not planted in ~ 28July
2012. 28 July ~
Establishment of nurseries began in the Direct seeding
beginning of June and peaked on 10 June.
Transplanting was carried out mostly from 15
to 22 July. The duration between the peak
time of nursery establishment and transplant-
ing was 5 weeks, while it was usually 30 days
in N Village. In 2011, peak times of nursery
establishment and transplanting occurred in
early June and early July respectively (Ikeura
et al., 2012). Transplanting in 2012 was de-
layed one or two weeks compared to 2011; the
delay in transplanting appeared to be a result Fig. 4 Completion time of transplanting
of little rain being received from 10 to 30 June
2012.
A map illustrating completion time of transplanting in each field is provided in Fig. 4. Trans-
planting was started earlier in Area A than in Area B. It was expected that transplanting would be
completed earlier at the upstream fields for which water could be taken from canals directly. How-
ever, transplanting was delayed in some upstream fields despite their location near canals. Trans-
planting in three upstream fields was completed after 4 August. The four fields that were transplanted
the latest were owned by farmers in the adjacent village who might have other fields in their own vil-
lage. In N Village, a delay in transplanting was caused by competition for labour in 2010 and 2011
(Ikeura et al., 2012). In addition to water conditions and labour competition, farmers’ prioritization
of their own fields may have been a factor in determining the time of transplanting.
Table 2 Rice yield in the sampling fields

Yield
Std. dv. Note
Field No (ton/ha)
Field 2 2.41 1.10 a
Field 5 4.58 0.93
Field 6 3.77 1.37
Field 7 4.31 1.31
Field 9 3.61 1.44
Field 14 3.78 0.67
Field 19 4.60 1.43
Field 22 4.76 0.97
Field 24 3.10 0.27
Field 25 4.83 1.12
Field 26 4.95 0.66
Field 27 3.14 0.50
Field 28 3.95 1.24
Field 29 4.78 1.21
Field 30 3.96 0.40
Field 31 4.73 0.90
Field 32 3.59 0.29
Field 33 3.18 0.75
Field 34 4.33 0.76
Field 35 4.68 1.65
Field 37 4.08 0.77
Field 38 2.99 1.35
Field 40 2.56 0.74 b
Field 41 4.08 0.68
Field 43 3.42 1.80
Field 44 3.89 1.90
Field 45 3.11 0.93
Field 46-1 5.14 0.84
Field 46-2 3.44 1.09
Field 51 6.80 3.53 a, b
Average yield of 3 replications
a, b: significant difference with alpha<0.05
Yield was calibrated as moisture content =14%
3.3 Rice yield

Table 2 and Fig. 5 show the average rice yield of 3 replicates in each sampling field. Average
yield for the entire survey area was 4.02 t/ha. The minimum yield was 2.41 t/ha, at Field 2, and the
maximum yield was 6.80 t/ha, at Field 51. Tukey–Kramer HSD tests indicated that there were sig-
nificant differences (p < 0.05) between Fields 2 and 51 and between Fields 40 and 51. Field 2 was
located in the upper part of Area B; however, water supply was not stable and the field dried out fre-
quently. Field 40 was located in the lowest part of the Area A; there was water shortage when it did
not rain, and flooding occurred after heavy rains. Flooded conditions generally cause damage to rice
and decrease yield. Minagawa et al. (2013) reported that rice yield was reduced by flooding at boot-
ing stage, especially under long-term, complete submergence or submerged conditions lasting more
than 3 days. In low-elevation fields, such
situations might damage rice growth and 6 ton <
productivity. Field 51 was under watery 5 ton <
conditions, because it was located near a 4 ton <
swamp. A local, long-stem rice variety 3 ton <
(Makkua) was planted in this field. It was
2 ton <
suggested that application of a suitable va-
riety and good water conditions contribute
to increasing yield. Highly productive fields
were located not only near the canal, but
also in plot-to-plot fields. Some of upstream
fields showed low yields. Table 3 shows
the average rice yield in the upstream and
plot-to-plot fields. The average yield in
plot-to-plot fields was larger than that in
Fig. 5 Map of rice yield in the survey area
upstream fields but the difference was not
significant.
Table 4 and Fig. 6 show average rice yield in two areas and cumulative discharge from two riv-
ers from 29 June to 15 November 2012. The discharge from River A was 95,700 m3, and it was
stably supplied to the fields in Area A throughout the planting period. In contrast, River B supplied
only 2,300 m3 of water during the period and it stopped flowing several times. Although the water
supply was different in the two areas, there was no significant difference in yield between Areas A and
B.
Table 5 shows the yield of each rice variety. Seven varieties of rice were planted in the sam-
pling fields and they were distributed in the areas. The varieties planted in only one field were clas-
sified as ‘the others’. Average yields were not significantly different and varied from 3.8 t/ha to 4.4
t/ha.
Table 3 Average rice yield in upstream fields
and plot to plot fields
Average
Std. dv. Note
Field type (ton/ha)
Upstream fields 3.74 1.20 n=33
Plot to plot fields 4.17 1.43 n=57
Nosignificant difference by t-test
Table 4 Average rice yield in Area A and Area B

Average
Std. dv. Note
Area (ton/ha)
Area A 4.09 1.39 n=72
Area B 3.74 1.20 n=18
Nosignificant difference by t-test
Fig. 6 Water discharge from River A and B
Table 6 Average rice yield of difference time of
Table 5 Average rice yield of each variety of rice transplanting
Average Average
Std. dv. Note Std. dv. Note
Variety (ton/ha) Translanting (ton/ha)
Tia khao 4.37 1.13 n=18 2 – 8 July 2013 5.00 2.51 n=9
Takkied 4.02 1.39 n=15 9 – 15 July 203 3.97 1.39 n=51
Tia daeng 3.83 1.17 n=33 16 – 22 July 2013 3.94 1.17 n=21
The others 4.02 1.72 n=24 Later than 23 July 3.51 1.72 n=9
Nosignificant difference by Tukey-Kramer HSD test Nosignificant difference by Tukey-Kramer HSD test
Table 6 shows average yield at the different transplanting times for all of the survey fields.
Although it also did not differ significantly, there was a decreasing tendency in yield at late trans-
planting times. Especially, one of the low-productivity field was transplanted at the end of July.
The time of transplanting appears to be one of the factors explaining the differences in rice yield in N
Village.
4. Conclusion
The factors behind the differences in yield were considered in terms of field location, water
supply, variety of rice, and time of transplanting. Except for the yields in two low-productivity fields
and one high-productivity field, no significant differences were observed in the yields from the survey
area. A wide difference was noted in the yields of each field, which was one of the reasons for the
lack of significant differences in yield among the different field conditions. The negative effects on
rice growth and yield seemed to be because of other factors or a combination of factors. In order to
clearly understand the effect of the issue described above, other influential factors such as soil fertility
and water quality should also be considered.
In 2012, although the survey area received little rainfall prior to transplanting, 960 mm of pre-
cipitation was received in August and September, which is almost the same amount as that received
during a typical year. Rice yield may be reduced in drought years; however, further confirmation is
necessary. Regarding time of transplanting, the yield was reduced at late transplanting times. Si-
paseuth et al. (2000) reported that yield of lowland rice increased with early transplanting and de-
creased by 30–60% when transplanting was delayed from the optimal time. This means that produc-
tivity may be increased by early transplanting in N Village. Thus, it is important to identify the best
time for transplanting to obtain the maximum yield.
References
FAO, 2011, Global Statistics Service - Food Security Indicators, Lao Peoples Domestic Republic,
http://www.fao.org/fileadmin/templates/ess/documents/food_security_statistics/monitoring_prog
ress_by_country_2003-2005/LaoPDR_e.pdf, Accessed 20 September 2012.
H. Ikeura, Hiroshi Ikeura, S. Inkhamseng, S. Vongphachanh, P. Xaypanya, 2012, Influential factors in
determining the timing of transplanting lowland rice, case study in Lao PDR, PAWEES 2012 In-
ternational Conference, 27-29 November 2012, Thailand
H. Ikeura, S. Vongphachanh, S. Phonchanmixay, K. Sisopha, 2012, The Subjects on Water Use to In-
crease Lowland Rice Yield in Lao PDR, Proceedings of Congress of The Japanese Society of Ir-
rigation, Drainage and Rural Engineering, 298-299 (In Japanese)
H. Minagawa, T. Masumoto, T. Yoshida, R. Kudo, I. Kitagawa, C. Zukemura, 2013, A Pseu-
do-flooding Experiment under Real Inundation Conditions by Using Clean and Turbid Water
Plots to Formulate Reduction Scales in Rice Yield - Design of the Experimental Plots and Inves-
tigation of Growth Conditions -, Technical report of the National Institute for Rural Engineering,
214, 111-121
Sipaseuth, P. Inthapanya, P. Siyavong, V. Sihathep, M. Chanphengsay, J. M. Schiller, B. Linquist and
S. Fukai, 2000, Agronomic Practice for Improving Yields of Rainfed Lowland Rice in Laos,
Increacing lowland rice production in the Mekong region: Proceedings of an International
Workshop held in Vientiane, 101, 31-40 (In Japanese)
WFP, 2007, Lao PDR, Comprehensive Food Security & Vulnerability Analysis, 34-35
THE FUTURE
D-08
Investigation of Organic Fertilizer to Reduce Insecticide

-Assessment of Paddy Ecosystem using Emergence Husks of
Red-dragonflies-
AODA Tadao*, KATANO Kai**, TOYAMA Kazunari**, JINGUJI Hiroshi***
We assessed the function of the environmental creativity of organic fertilizer. We selected

red-dragonflies as an indicator of paddy ecosystems. In this study we focused on three species of
red-dragonflies, i.e. Sympetrum frequens, Sympetrum darwinianum and Sympetrum infuscatum. We
analyzed relationship among number of habitation, pesticide (insecticide and herbicide) usage, water
management and soil physical condition since 2011 at the paddy fields in Niigata Prefecture, Japan.
Here we replaced number of emergence husks as habitation of dragonflies. We collected emergence
husks under the cooperation with farmers. The numbers of tested fields are 72 in 2011 and 52 in 2012.
Consequently, we found that, 1) peak period of dragonfly emerges was at the last week of June, 2)
Sympetrum frequens is the major species occupied around 80% in Niigata Prefecture, 3) application
of insecticide for seed-box and mid-season drainage had negative impact to dragonflies’ life, 4)
insecticide of Neonicotinoid type was the worst effect for red-dragonflies’ habitat, and 5) number of
emergence husks of dragonfly at organic paddy field was not so much than we expected. Further
studies should be done to define the effect of water management, organic fertilizer and pesticide into
dragonflies’ habitat.
1. Introduction
The landscape with a group of red-dragonflies had been original scenery at rural area in Japan.
However, red-dragonflies decrease rapidly in rural area in these days. Jinguji, et al. (2009) and
Nakanishi, et.al. (2009) indicated that the reclamation of well-drained paddy field, excess usage of
pesticide (herbicide, insecticide and fungicide), and severe drainage in mid-season have negative
impact into the ecosystem of paddy field. However, we do not know the clear and accurate relation
among decreasing of dragonfly, water management and pesticide usage. On the other hand, Taira
(2012a and 2012b) indicated that neonicotinoid type insecticide damaged to human health. Also we do
not know time series of population of dragonflies for the past decades. Therefore we investigate the
relationship among pesticide usage, water management, and the number of emergence husks of
red-dragonflies in Niigata Prefecture, Japan.
2. Ecology of the red-dragonfly
We focused on three species of red-dragonflies. Sympetrum frequens, Sympetrum darwinianum

and Sympetrum infuscatum are the target species. Those eggs laied at paddy fields in previous fall,
incubated at spring after water inlet. Larvae grow up with feeding zoobenthos and cast off their skin
around ten times, and emerge at rice stems in the period of June to in the morning. They mature during
summer, and laying eggs at wet paddy fields at autumn (Inoue and Tani, 2010).

We prescribed red dragonflies as an indicator species, and investigated paddy ecosystems from
the habit of them. Since, red dragonflies are the universal creatures and famous in Japan. Detail
investigate procedures are as follows; 1) counting the inhabit dragonflies from the number of husks, 2)
clarifying the relationship among pesticide management, water management and dragonflies’ habit.
Where, water management defines water inlet at spring, midterm drainage at rainy season, on-off
irrigation at summer, and drainage before harvest. The midterm drainage interpreted to drain surface
water at paddy fields around one week almost thirty days after transplant. The aims of midterm
drainage are controlling offshoot, enhance root activity by the reduction of hydrogen sulfate and
promoting soil hardening. Hence midterm drainage is deeply related with life of dragonflies, since
have been done exactly same time at the emergence of them.
We classify rice producing methods by the pesticide and/or fertilizer usage. Conventional rice
production uses chemical fertilizer and pesticide. Organic rice farming is the procedure without any
pesticide or chemical fertilizer. Reduction rice farming is the method to reduce the amount of
pesticide.
3.1 Investigation of Emergence husks of red-dragonfly

3.1.1 Collection and identification of emergence husks
The test paddy fields are located in Niigata Prefecture, Japan. The number of test paddy fields
was 72 in 2011 and 51 in 2012. Control was 15 paddy fields located in Miyagi Prefecture, Japan, in
2010. Because of the comparative investigation, we collect emergence husks from the paddy fields in
relatively low and level in topography. We asked to collect emergence husks to rice farmers from 12th
June to 16th July, 2011. The collecting area of emergence husks was three-row rice stubbles along
shorter border in outlet side (see Figure 1). Collected husks were packed in plastic cases to keep their
shape, and posted to the laboratory in Niigata University. The points to identify the species dragonfly
larvae were the size of emergence husk, shape of lower lips, and length of side spines at 8th node and
9th one (see Figure 2). After identification of dragonfly species, we analyzed relationship among
number of husks, water management and pesticide usage.
inlet
drainage channel
→irrigation channel
outlet
Figure 1 Sampling points of emergence husks at test paddy fields in 2012.

We collected emergence husks on the rice stubbles near the shorter border in
side of drainage channel.
3.1.2 Questionnaire of rice farming procedure
Lower lip
8th node
9th node Side spine
Figure 2 Points to identify dragonfly husks

Side spine of 8th node straight toward 80% of 9th node.
We did questionnaire to farmers to clarify the relationship between dragonfly habitat and rice
farming procedure. The question items were as follows; 1) address of farmers and field, 2) water
management (date of tillage, water inlet and transplant), 3) seed box insecticide (product name and
amount of usage), 4) herbicide (product name and amount of usage), 5) weed control method (rice
bran, rubbish soybean, paper multi, hybrid from a wild and domestic duck, and so on), 6) insecticide,
except of rice seedling (product name and amount of usage), 7) mid-term drainage (date of starting
and final date), 8) degree of interest (degree of burden, reason of participation, degree of
understanding and impression of participation). We identified the number of the ingredients of seed
box insecticide and herbicide from product names.
4. Results and Discussions
4.1 Collection of emergence husks

4.1.1 Time series of emergence of red dragonflies
We showed time series of dragonfly emergence in Niigata Prefecture in 2011 in Figure 3. The
peak period of emergence of red dragonflies was the last week of June. And they had the same trend in
2012 in Niigata Prefecture. The other hand, the peak period of emergence of red dragonflies in Miyagi
Prefecture was first week of July, which was one week later than Niigata.
300
250
Sympetrum
アキアカネ frequens
Number of emergence
Sympetrum
ノシメトンボ darwinianum
200
Sympetrum
ナツアカネ infuscatum
総羽化殻数
150 Except Sympetrum genus

アカネ属以外
Undistinguishable
不明
100
50
0
6/12～6/18 6/19～6/25 6/26～7/2 7/3～7/9 7/10～7/16
Study period
調査期間
Figure 3 Time series of dragonflies’ emergence in Niigata Prefecture, 2011
4.1.2 Occupied species of red-dragonfly
The number of emergence husks cooperated with farmers were 536 in 2011 and 1768 in 2012.
A breakdown in the manifestation rate of species in 2011 was as follows; Sympetrum frequens was
82%, Sympetrum darwinianum was 7% and Sympetrum infuscatum was 3%, Non-Sympetrum genus
was 3% and undistinguishable was 5% (see Figure 4). In 2012, the rate of species was; Sympetrum
frequens 58%, Sympetrum darwinianum 34% and Sympetrum infuscatum 7% and Non-Sympetrum
genus 0%. Therefore, Sympetrum frequens defines as the occupied species in Niigata Prefecture. On
the otherhand, the manifastation rate of red-dragonfly in Miyagi Prefecture was different from its in
Niigata Prefecture. The details was; Sympetrum frequens was 47%, Sympetrum darwinianum was 28%
and Sympetrum infuscatum 25%. Geographyical and climatic difference might affect habit of
red-dragonfly. However, we are still under investigation what makes geographical difference in the
habit of red-dragonfly.
Sympetrum frequens
Sympetrum darwinianum
Sympetrum infuscatum
Non-Sympetrum genus
Undistinguishable
Figure 4 Manifestation rate of emergence husks of red-dragonflies in

Niigata Prefecture, 2011.
4.1.3 Midterm drainage and emergence of red-dragonfly

We showed the Effect of midterm drainage into emergence of red-dragonfly at Niigata
Prefecture, 2011 in Figure 5. Though, paddy fields drained before emergence have small number of
husks, paddy fields drained after emergence or no-drainage has relatively large number of them. We
sawed this trend in 2012, too. Since dragonfly larvae vulnerable to dry, midterm drainage affects
negatively into emergence.
We hypothesized that the earlier inlet water into paddy field, the larger number of emergence
husks. However we did not have clear relation between the date of water inlet and number of husks.
16
14
平均羽化殻数（標準偏差）
Number of emergence
(standard deviation)
12
10
8
6
4
2
0
中干し無し
Non-midterm drainage 中干し有り
Midterm drainage
Figure 5 Effect of midterm drainage to emergence of red-dragonflies in
Niigata Prefecture, 2011
4.1.4 Effect of herbicide and/or seed-box insecticide to dragonfly emergence
We showed effect of herbicide and/or seed-box insecticide to dragonfly emergence in 2011 in
Figure 6. The number of emergence husks at the paddy fields applied only herbicide was larger than
organic paddy fields. The number of emergence husks at conventional paddy fields applied herbicide
and seed-box insecticide was lowest. We confirmed these trends in 2012 in Niigata Prefecture and in
Miyagi Prefecture. Therefore, we do not declare that the herbicide affects negatively into life of
dragonflies. We found that seed-box insecticide have negative impact to dragonflies habits.
18
16
Number of emergence
平均羽化殻数(標準偏差)
14
12
10
8
6
4
2
0
除草剤・箱施用剤
Seedbox 有機
Organic
insecticide 除草剤のみ
Herbicide
+ Herbicide
Figure 6 Effect of herbicide and/or seedbox insecticide to dragonfly emergence
in Niigata Prefecture, 2011.
We showed the relationship between branch of seed-box insecticide and number of dragonfly
emergence in Figure 7. Seed-box insecticide, used in Niigata Prefecture, were classified mainly in
three branches as follows; Neonicotinoid type (included fipronil), Hetero type, and Nereistoxin type.
Though the sample sise is not large enough, we do not have obvious conclusion. Number of
emergence husks of red-dragonfly in conventional paddy fields applied neonicotinoid type insecticide
was low conspiciously. We supposed residual effect of neonicotinoid impacts negatively on the
ecology of dragonfly.
20
Number of emergence
18
16
14
平
均 12
羽 10
化
殻 8
数
6
4
2
0
ヘテロ系ネライストキシン系ネオニコチノイド系
Neonicotinoid Hetero （フィプロニル含む）
Nereistoxin
N=8 N=3 N=9
Figure 7 Relationship between branches of the seed-box insecticide and

dragonfly emergence in Niigata Prefecture, 2012.
5. Conclusion
The occupied species of red dragonflies were Sympetrum frequens in Niigata Prefecture, Japan.
The peak period of emergence was the last week of June in Niigata Prefecture, and was the first week
of July in Miyagi prefecture, Japan. Midterm drainage before emergence and seed-box insecticide
impacted negatively on habit of dragonfly. Insecticide based on neoticotinoid was worst effects to
dragonfly emergence.
Midterm drainage is the water management to enhance rice quality and quantity, and to improve
the trafficability of combine at harvest. On the other hand, some rice farmers have been studying
producing rice without midterm drainage to improve their paddy environment, i.e., lower the density
of transplant of rice. And most rice farmers carry out midterm drainage at the peak period of
emergence of dragonfly at the end of June. In these days, agricultural administration needs to conserve
not only rice productivity but also biodiversity. As a countermeasure, one to drain surface water at
midsummer after the emergence of dragonfly or to reduce amount of water.
Seed-box insecticide extracted from rice root, translocated to stem and leaf, and remained
relatively long tome in rice plant. So rice farmers require sensitivity more than before when they used
seed-box insecticide to protect rice productivity and biodiversity. In particular, precautionary principle
should be adopted in use of neonicotinoid type insecticide. Since, Taira (2012a and 2012b) pointed out
insecticide based neonicotinoid damage on human health.
Only few studies had been focused on the relationship between the ecology of red-dragonfly
and rice farming. We studied the present condition of paddy fields’ environment used red-dragonfly as
an indicator. Unfortunately we did not have enough number of samples fields in conventional way to
determine the relationship between red-dragonfly and pesticide management in 2011 and in 2012.
Further study should be necessarily to have more detailed evaluation with increasing sample fields. To
establish sustainable paddy-rice cultivation harmonized with nature, one should enhance
environmental awareness to local residents.
5.1 Acknowledgement
This work was financially supported by JSPS KAKENHI Grant Number 23405045, Strategic
Study Support from Japanese Society of Irrigation, Drainage and Reclamation Engineering;
Encouragement of Scientific Study on Biodiversity, Sado-City, Niigata Prefecture, Study fund of
Sasakami Food and Agricultural Network. The authors appreciate to collect the emergence husks of
dragonflies by Promotion Network of Organic Farming in Niigata Prefecture, Society of Food Support
in Niigata and Satoyama Promotion Council to Live with Ibis (Nipponia Nippon).
6. References
Inoue, K. and Tani, K. (2010). All about Red Dragonflies. Dragonfly Publishing Co. 34-175.
Jinguji, H. Tashiro, S. Sato, T., Tsuyuzaki, H., and Kondo, T. (2006). “Effect of Cultivation Methods
in a Controlled Mixing Tillage of Plow Layer on Habitat Condition of the genus Symptrum.”
Trans. JSIDRE, 241. 133-140.
Jinguji, H., Uéda, T., Goka, K., Hidaka, K., and Matsura, T. (2009). “Effect of Imidacroprid and
Fipronil insecticide Application on the Larvae and Adults of Sympetrum frequens (Libellulidae:
Odanata).” Trans JSIDRE, 259, 35-41.
Jinguji H., Thuyet, D. Q., Uéda, T., and Watanabe, H. (2012). “Effect of imidacloprid and fipronil
pesticide application on Sympetrum infuscatum (Libellulidae: Odanata) larvae and adults” Paddy
and Water Environment, DOI 10.1007/s10333-012-0317-3.
Nakanishi, K., Tawa,K., Kanbara,B. Noma, N. and Sawada,H.(2009). “Comparison of macro-aquatic
communities among paddy fields under different cultivation management systems” Jpn. J.
Environ. Entomol. Zool. 20(3), 103-114.
Taira, K. (2012a). “Health effects of neonicotinoid insecticides－Part 1: Physicochemical
Characteristics and Case Reports－” Japanese Journal of Clinical Ecology 21(1), 24-34.
Taira, K. (2012b). “Health effects of neonicotinoid insecticides－Part 2: Pharmacology, Application,
Regulation, and Discussion－” Japanese Journal of Clinical Ecology 21(1), 35-45.
Suzuki, N. (2006). Effect of Cultivation Management of Paddy Fields on Biota and Soil Physical
Conditions. Bachelor thesis. Faculty of Agriculture, Niigata University, 1-30.
Session 5
THE FUTURE
E-01
Screening Rice (Oryza sativa L.) Varieties Suitable for System of

Rice Intensification (SRI)
K. Noborio*, J. Lanceras-Siangliw**, K. Katano*, M. Mizoguchi***, T. Toojinda**

*School of Agriculture, Meiji University, Kawasaki, Japan
**Rice Gene Discovery Unit, BIOTEC, Kasetsart University, Kamphaeng Saen, Nakhon
Pathom, Thailand
***Graduate School of Agriculture and Life Sciences, University of Tokyo, Tokyo, Japan
ABSTRACT
Although system of rice intensification (SRI) has been popular in Southeast Asia, some workers
claimed that there were no increases in yield. We thought that it might be attributed to less exploration
of rice (Oryza sativa L.) varieties adapted with SRI. In rice paddy fields at Kasetsart University in
Kamphaen Saen, Thailand, 169 rice varieties and 5 control varieties were planted using augmented
design under aerobic condition for SRI and under flooded condition or anaerobic condition using
conventional cultivating method. Soil moisture content ranged between 17 to 25 % while redox
potential was between 300 and 500 mV after the first measurement, which was 100 mV. This can be an
indication that the field was maintained aerobic throughout the growing period. Some varieties such
as IR70 and IRGC35747 provided more yield in SRI or aerobic condition than in flooded condition
although some showed fewer yields in SRI. Our results implied that appropriate varieties should be
used for SRI in region to region.
1. Introduction
System of rice intensification (SRI) was introduced by Father Laulanié in Madagascar in 1980’s
(Laulanié, 2011). He claimed that local as well as improved varieties of rice (Oryza sativa L.)
increased yields with SRI. Review articles for SRI practices are found elsewhere such as Uphoff et al.
(2011). Ceesay et al. (2006) reported that 2-3 times yield increases in three indica varieties of rice with
the SRI practice in tropical Gambia. The SRI practice increased yield by 22-23% and 44-67% for
japonica varieties of rice in sub-tropical Hangzhou, China (Zhao et al., 2009) and in temperate
Tsukuba, Japan (Minamikawa and Sakai, 2005), respectively.
In temperate regions of monsoon Asia, however, workers reported yield reduction in japonica
and indica varieties of rice with SRI practices. Chapagain and Yamaji (2010) and Kudo et al. (2012)
respectively reported 8% and 13-18% reduced yield for japonica varieties of rice with SRI practices
near Tokyo, Japan. Choi et al. (2013) also showed that there was 8-24% lesser yield for a local
japonica variety with SRI practices than that with the conventional flooded practice in Korea. It was
also reported even for an indica variety of rice SRI practices decreased yields near Tokyo, Japan
(personal communication with N. Shimoozono, 2013).
Water management practices might be very critical for increasing yields especially in temperate
regions (Kudo et al., 2013). We also speculate that varieties of rice may largely be attributed to
varying yields with SRI practices. In this paper, we report preliminary results of rice yields varying
from variety to variety with SRI practices.
In rice paddy fields at Kasetsart University in Kamphaen Saen, Thailand, 169 rice varieties and
5 indica varieties bred for flooded cultivation as control were planted using augmented design under
aerobic condition for SRI and under flooded condition or anaerobic condition using conventional
cultivation. For aerobic and flooded conditions all the varieties were seeded on June 7, 2012. For the
aerobic condition, 15-day old seedlings were transplanted on June 22, 2012 whereas 30-day old
seedlings were transplanted for the flooded condition on 15 days later than the aerobic condition.
Water was applied only twice in July and after that the field relied on rainfall. Redox potential and soil
moisture were measured from July to August because of continuous rain in September until flowering
in October (Fig. 1).
120
100
Rainfall (mm)
80
60
40
20
0
600
500
Redox (Mv)
400
300
200
100
26
24
Soil Moisture (%)
22
20
18
16
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160
June July Aug. Sept. Oct. Nov.
Flowering time
Germination Water Water Harvesting

Application Application
Transplanting
Fig. 1. Environmental condition in the aerobic field during the growing period in wet season 2012 at
Kasetsart University, Kamphaeng Saen, Thailand.
Environmental conditions for the growing period are shown in Fig. 1. Soil moisture content
ranged between 17 to 25 % while redox potential was between 300 and 500 mV after the first
measurement, which was 100 mV. This may be an indication that the field was maintained aerobic
throughout the growing period for the SRI practice. The anaerobic field was flooded throughout the
growing period for the conventional practice (data not shown). The SRI and flooded fields are shown
in Picture 1. Soil cracks appeared in the SRI field were as wide as 3cm on the soil surface. Healthy
roots were visible through the cracks.
Picture 1. The SRI (left) and flooded (right) fields on August 2, 2012.
600.0
SRI (Aerobic)
500.0 Flooded (Anaerobic)
400.0
300.0
200.0
100.0
0.0
Ketan Lumbu IRGC35747 ( …
CHAINT1
Pitsanulok
Suphanburi 60
SPTC99039
SPTC04069
RD15
PTT1
SPTC99052 ( PTT )
SPTC04072 ( PTT )
SPTC04070 ( PTT )
SURIN1
Pinkaset 3
IR29
IR40
IR42
IR70
IR72
Sintanur (SRI)
Bueogi ( PTT )
FL496
PSBRc9
IR57514
PSBRc30
PSBRc70
WS48 PTT1
WS48 SPR2
WS48 SPR90
RGDU 07345-4-24-1
RGDU 07345-4-24-41
RGDU 07336-43-10-24
PSL 00034-50-2-1-8
PSL 00034-50-2-1-3
PSL 000504-54-1-2-2
Fig. 2. Grain yield (kg/ha) of selected rice varieties under SRI (aerobic) (blue) and flooded (red) condition.
Grain yields varied with varieties of rice as shown in Fig. 2. All the 5 control varieties
(CHAINT1, PTT1, IR57514, SURIN1, and Suphanburi60) decreased yields in the SRI condition as
expected. Some varieties such as PSBRc9, SPTC99052, WS48 SPR90, and Pitsanulok did not differ in
the SRI and the flooded conditions. Some varieties such as IR70, IRGC35747, and Bueogi increased
yields under the SRI condition. Those varieties may be more suitable than others for SRI practices in a
tropical region. Similar screening experiments should be conducted in other climatic conditions.
Acknowledgements
This work was partly supported by Grant-in-Aid for Scientific Research (A) (No. 23255014, PI: M.
Mizoguchi) by JSPS and Program for Establishing Strategic Research Foundations in Private
Universities (No. S0901028, PI: K. Noborio) by MEXT of Japan. We would like to acknowledge Mr.
Sakchai Suakham for his assistance in the field management in Thailand.
References
Ceesay, M., Reid, W.S., Fernandes, E.C.M., and Uphoff, N.T., 2006, “The effects of repeated soil
wetting and drying on lowland rice yield with system of rice intensification (SRI) methods.”
International Journal of Agricultural Sustainability, 4, 5-14.
Chapagain, T., and Yamaji, E., 2010, “The effects of irrigation method, age of seedling and
spacing on crop perform ance,productivity and w ater-wise rice production in Japan.” Paddy and
Water Environment, 8, 81-90.
Choi, J.-D., Park, W.-J., Park, K.-W., and Lim, K.-J., 2013, “Feasibility of SRI methods for reduction
of irrigation and NPS pollution in Korea.” Paddy and Water Environment, 11, 241-248.
Kudo, Y., Noborio, K., Kato, T., and Shimoozono, N., 2012, “Effects of intermittent irrigation with
different intervals on greenhouse gas emissions and rice yield.” Transactions of the Japanese
Society of Irrigation, Drainage and Rural Engineering, 80, 507-514. (in Japanese with English
abstract)
Kudo, Y., Noborio, K., Shimoozono, N., and Kurihara, R., 2013, “The best water management
practice for mitigating greenhouse gas emissions and maintaining rice yield in paddy fields.”
Agriculture, Ecosystems and Environment (in review)
Laulanié, H., 2011, “Intensive rice farming in Madagascar.” Tropicultura, 29, 183-187.
Minamikawa, K., and Sakai, N., 2005, “The effect of water management based on soil redox potential
on methane emission from two kinds of paddy soils in Japan.” Agriculture, Ecosystems and
Environment, 106, 397-407.
Uphoff, N., Kassam, A., and Harwood, R., 2011, “SRI as a methodology for raising crop and water
productivity: productive adaptations in rice agronomy and irrigation water management.” Paddy
and Water Environment, 9, 3-11.
Zhao, L., Wu, L., Li, Y., Lu, X., Zhu, D., and Uphoff, N., 2009, “Influence of the system of rice
intensification on rice yield and nitrogen and water use efficiency with different n application
rates.” Experimental Agriculture, 45, 275-286.
THE FUTURE
E-02
Effect of SRI Methods on Water Use, NPS Pollution Discharge,

and GHG Emission in Korean Trials pan
Joongdae Choi*, Gunyeob Kim**, Woonji Park***, Suin Lee****, Deogbae Lee*****
and Dongkoun Yun******
*Professor, Department of Regional Infrastructures Engineering, Kangwon National U
niversity, Chuncheon, Gangwondo, South Korea; jdchoi@kangwon.ac.kr
**Researcher, Agro-climatology Lab., Rural Development Administration, Suwon,
Gyeonggido, South Korea; gykim@rda.go.kr
***Post-doctoral researcher, Department of Regional Infrastructures Engineering, Kan
gwon National University, Chuncheon, Gangwondo, South Korea; parkwoonji@hanmai
l.net
****Researcher, Department of Regional Infrastructures Engineering, Kangwon Nation
al University, Chuncheon, Gangwondo, South Korea; tndls0915@hanmail.net
*****Director, Agro-climatology Lab., Rural Development Administration, Suwon, Gy
eonggido, South Korea; ledb419@korea.kr
******Researcher, Research Group on Rural Community Development, Rural Researc
h Institute, Ansan, Gyeonggido, South Korea; ydkibm@ekr.or.kr
*Corresponding author: jdchoi@kangwon.ac.kr, ☎+82-33-250-6464, (fax)+82-33-251-1
518
Abstract
An experimental plot study was performed in 2011 to measure the effect of paddy irrigation
management in Korea on rice yield, water use, non-point source (NPS) pollution discharge, and
greenhouse gas (GHG) emissions. A locally-bred Japonica rice cultivar was cultivated in 12 plots
sized 5 x 15 m. Experimental treatments were conventional paddy cultivation (CT), conventional
paddy cultivation with SRI water management (CS), conventional cultivation with two forced mid-
season drainages (CD), and SRI methods with transplant spacing between hills of 30x30 cm (SRI-30)
and 40x40 cm (SRI-40). Each treatment was replicated. The soil texture was a loam with organic
matter content of 28±1 g/kg. Evapotranspiration consumed most of the water in the plots. Irrigation
water supplied was reduced by 49.4% and 47.6% in the SRI and CS treatments, respectively, compared
with the CT treatment. Reductions in NPS pollution load in the water with SRI management ranged
from 16.5 to 53.9% in the CS plots and from 27.1 to 46.0% in the SRI plots. If CH4 and N2O from the
CD and CS plots are converted to CO2 equivalent, GHG emissions were reduced by 65.5% and 72.8%,
respectively, compared with emissions from the CT plots. The polished rice yield from the SRI-30-3
(6.47 ton/ha) and CS (6.34 ton/ha) plots was increased by 20% and 17%, respectively, over the yield
from the CT plots. These trial results indicate that SRI water management in Korean paddy farming
could significantly decrease rice crop irrigation water requirements and also discharges of NPS water
pollution from the paddy fields and GHG emissions into the atmosphere.
Keywords: SRI method, NPS pollution, greenhouse gas, rice yield, irrigation requirement
1. Introduction
Modern agriculture must increase its productivity and consider environmental sustainability by
developing and adapting environmentally-friendly agricultural technologies. These technologies
should improve agricultural productivity, reduce crop water consumption, ensure water quality and
sustain natural ecological systems, and reduce greenhouse gas (GHG) emissions to cope with the
effects of climate change. Paddy fields cover more than 60% of Korea’s arable land and consume
about 55% of the total water used. It is well recognized that non-point source (NPS) pollution from
paddy fields has a demonstrable effect on the water quality of the country's rivers and lakes due to
their size and large water consumption. Reducing the amount of freshwater devoted to irrigation
supply, particularly for paddy production, is an important part of any long-term national water supply
strategy and is required to improve water quality on a national scale. Reducing NPS pollution in the
paddy fields is especially important to achieve the target water quality of public waters under the total
maximum daily load (TMDL) policy in Korea.
The Korean government has worked hard to reduce the country's GHG emissions by developing
technologies and policies for the mitigation and retardation of global warming. The target reduction of
GHG emissions from the agriculture sector, including the fishery and forestry sectors, has been set to
5.2% of the total target reduction. Because livestock and rice paddy are leading contributors of GHG
emissions in agriculture (FAO, 2011), a reduction of GHG emissions in Korean paddy fields would
play an important role in meeting the reduction target. New methods or technologies are required in
Korean flooded rice farming to improve rice productivity while reducing irrigation water consumption,
NPS pollution discharges, and GHG emissions.
The system of rice intensification (SRI) is reported to successfully meet these requirements.
Studies on SRI in the past decade have shown that it could save 25- 40% or more of the water
requirement for irrigated rice production(Dixit, 2005; Hameed et al., 2011; Nyamai et al., 2012;
Krupnik et al., 2012); increase rice productivity by 50-100% and even more in some parts of
developing countries (Hameed et al., 2011; Thakur et al., 2010; Thakur et al., 2011); and reduce the
use of labor, fertilizers and pesticides (Chapagain and Yamaji, 2010; Uphoff et al., 2011). Most studies
on the SRI thus far have measured and described the relationships among rice yields, water saving,
seedling ages, transplant spacing, organic compost uses, weeding methods, and labor cost reductions.
The environmental effects of SRI practices on water quality and GHG emissions have not been
systematically investigated to an appropriate extent.
To assess what, if any, effects SRI management can have on soil, air and water quality, studies
are needed that measure the relationships among SRI methods, NPS pollution discharges, and GHG
emissions. This study undertook to experimentally measure water-saving levels as well as any
reductions of NPS pollution discharges and GHG emissions in Korean experimental field plots where
SRI methods were applied.
No definitive conclusions can be drawn from a single study of phenomena that are so complex
and contingent. However, the results from the trials reported here should encourage further efforts to
evaluate these relationships in more contexts. If confirmed, the research results reported here could
contribute to the development of more successful national policies for improving water quality and
mitigating global warming. If farmers can increase their production and income by changing their
current rice cultivation practices, they will have economic incentives to manage their water in ways
that are more beneficial for the natural environment.
Twelve experimental plots sized 5x15 m each were constructed in an existing paddy field on an
experimental farm at Kangwon National University in northeastern Korea in 2010. The plots had been
used to cultivate a variety of rice under SRI water management during the 2010 growing season (Choi
et al., 2013), and those SRI experiences permitted better application of SRI methods in 2011. The data
presented in this paper were collected during the 2011 growing season. A locally-bred Japonica rice
variety was cultivated, Odaebyeo (Oryza sativa L.) known for early maturation and cold tolerance.
Seeding, transplanting and harvesting were done on April 22, May 6 and September 6,
respectively. The experimental treatments implemented included conventional cultivation in terms of
seedling age, spacing and other practices described below (CT); conventional cultivation with SRI
water management (CS); conventional cultivation with two forced mid-season drainages (CD); and
SRI methods with transplant spacing of 30x30 cm (SRI-30) and 40x40 cm (SRI-40). Each treatment
was replicated.
The SRI cultivation had no continuous flooding as is conventionally done (SRI principles of
water management reduce and control water applications), wider spacing of single transplants, soil
aeration, and organic fertilizer usage. The CT, CD and CS plots were mechanically transplanted with
seedlings about 21days old, with spacing of 30x15 cm, transplanting between four and six seedlings
per hill. The SRI-30 and SRI-40 plots were manually transplanted with the same seedlings of 14-day
age. However, each SRI plot was divided into 3 sub-plots, and each sub-plot was transplanted with 1,
2, or 3 seedlings per hill, respectively. Where there were 2 or 3 seedlings per hill, the shape and
spacing between seedlings was about 7 cm as shown in Fig. 1 (Zheng et al., 2004). These sub-
treatments were named SRI-30-1, SRI-30-2, and SRI-30-3, and SRI-40-1, SRI-40-2, and SRI-40-3,
accordingly, to consider whether intra-hill spacing would have a beneficial effect as well as inter-hill
spacing.
Fig. 1 Modified transplant method of 2(a) and 3(b) seedlings per hill for the SRI plots
The preparation and application of seed germination, field plowing and leveling, and application
of chemical fertilizer and pesticide were the same for all of the treatments. In addition to the
recommended chemical fertilizer application (N 110 kg/ha, P2O5 45 kg/ha and K2O 57 kg/ha),
vermiculture compost (OM 375.8 g/kg, TN 2.72% and available phosphate 750 mg/kg) was applied at
about 1.1 ton/plot. The SRI plots were weeded manually, while the CT, CD and CS plots controlled
weeds using herbicides. The SRI principle of active soil aeration, by use of a manual weeder that
breaks up the soil surface to provide more oxygen to the roots and soil biota to promote the growth
and health of these factors in crop performance, was not evaluated in these trials.
The seedlings used for conventional transplanting were raised in plain plastic trays, and those
used for the SRI methods were raised in 200-port plastic trays. Local rice farming practices were
applied to the CT plots. The same was done for the CS plots, but their water management followed
SRI principles. The same was also done for the CD plots, but two forced mid-season drainages of
about 10 days each were made during the growing season before the plots came to panicle formation.
Irrigation pipes and flow meters were laid throughout the plots to measure the exact amount of
irrigated water supplied. A measuring flume and an automatic water level gauge were placed at the
outlet of each plot to measure the drainage amounts. The irrigation and drainage water samples were
collected either manually or using an automatic water sampler (6712 Portable Sampler, Teledyne
ISCO, USA) or Coshocton Wheel sampler (Bonta, 2002; Bonta and Pierson, 2003). Three transparent
plastic gas chambers per plot were placed in the middle of the CT, CS and CD plots to measure GHG
emission. Each chamber, sized of 60x60x150 cm, had a lid at the top that opened and closed and a
plastic access part where air samples were collected by a syringe. The methods of Yagi (1991) were
applied to take a 60-ml air sample twice a week between 10:00 and 15:00. The air, water and soil
temperatures and soil Eh in the chambers were also measured during the air sampling process.
The water and gas samples were analyzed according to the national guidelines (Ministry of
Environment, 2007). Suspended solids (SS), biological oxygen demand (BOD5), chemical oxygen
demand (CODCr and CODMn), total nitrogen (TN), and total phosphorus (TP) were analyzed from the
water samples. Methane (CH4) and nitrous oxide (N2O) were measured from the gas samples. The
common properties of the soil, infiltration and evapotranspiration were also analyzed and measured,
respectively, according to the relevant national standards (Ministry of Environment, 2009). The
collected data were analyzed in terms of the treatment, and required statistical analyses were also
performed.
The soil of the plots had the following composition: 48.6% sand, 33.8% silt and 17.5% clay,
classifying it as a loam. The organic matter content and pH of the soil were 28±1 g/kg and 6.3±0.2,
respectively. The exchangeable cations of Ca, Mg, and K were 6.3±0.3 cmol/kg, 1.9±0.2 cmol/kg, and
0.5±0.06 cmol/kg, respectively. Ca and Mg fell into the common range of typical Korean paddy soil
properties, while K was a little higher. The available phosphate and silicate were measured at 135±7
mg/kg and 163±23 mg/kg, respectively. The measured average infiltration and evapotranspiration of
the plots were 0.3 mm/day and 4.6 mm/day, respectively. This means that evapotranspiration was a
major mechanism consuming water in the paddy plots, and that the infiltration loss was insignificant.
During the growing season from May to September 2011, a total rainfall of 1,609.9 mm was
recorded at the experimental plots. There were 48 rainfall events sized between 0.3 and 232.6 mm.
Two large rainfall spells occurred June 22-28 and July 24-27, when 416 mm and 500 mm of rainfall
were recorded, respectively. Runoff ratios for CT and SRI plots ranged from 0.83 to 0.96 and 0.70 to
0.89, respectively. This means that the rainfall runoff from the SRI plots was smaller than that from
the CT plots, and hence that the discharge of NPS pollution from the SRI plots was lower than that
from CT plots, as the pollution is mostly transported with the runoff.
3.1 Irrigation requirement and water quality

Table 1 shows the irrigation water supply to the plots and the irrigation reduction rates of the
CD, CS and SRI plots compared with the irrigation water supplied to the CT plots. The amount of
irrigation supply mainly depended on the amount of the evapotranspiration and rainfall. The SRI and
CS treatments were shown to reduce irrigation supply by 49.4% and 47.6%, respectively, compared
with the CT treatment. The irrigation reduction level observed previously in 2010 was 55.6%. The
differences in the reduction rate between 2010 and 2011 were mainly attributable to the rainfall
differences between the years. It should be noticed that the puddling water was excluded when the
reduction rates were computed in Table 1. If it was included, the rate decreased slightly to 14.4%, 37.4%
and 47.2% for the CD, CS and SRI plots, respectively. In any cases, SRI water management was seen
to significantly reduce the irrigation requirement, and this helped reduce the NPS pollution discharges
from the paddy fields.
Table 1 Irrigation water supply to the plots and the irrigation reduction rate of the CD, CS and SRI plots
compared with that of the CT plots
Measurement Irrigation supply (㎥)

Interval CT CD CS SRI
5.10 –5.13 (puddling)* 16.5 14.5 15.0 9.7
5.14–5. 20 5.0 4.8 2.6 2.5
5.20–5. 31 7.2 7.4 3.9 3.7
6.1 –6.10 12.2 10.5 5.4 4.8
6.11 - 6.30 11.7 7.7 5.9 5.7
7.1-7.31 5.6 5.2 3.6 3.8
8.1-8.31 4.9 3.9 3.1 3.1
Total supply (㎥) 46.6 39.5 24.5 23.6
Reduction rate (%) - 15.2 47.4 49.4
* Puddling water excluded when the reduction rates were computed
Table 2 shows the measured water qualities of the irrigation water used during the 2011
growing season compared to Korea’s irrigation standard. Except for TP, the measured water qualities
were within the national standard and did not show large differences from the available national
averages. The TP concentration of the irrigation source was quite high in 2010 for unknown reasons.
The irrigation water was considered to be good quality for irrigation in general.
Table 2 Korean water quality standards for irrigation, and the water quality measures for the irrigation
water used in 2011
Index Standard National average This study (2011)
pH 6.0∼8.5 7.8 7.5 ± 0.3
BOD (mg/L) ≤8 n.a.* 2.3 ± 0.5
COD (mg/L) ≤8 4.5 5.3 ± 1.2
SS (mg/L) ≤100 n.a. 12.9 ± 7.3
DO (mg/L) ≥2 9.4 8.6 ± 0.4
T-N (mg/L) n.a. 2.269 1.93 ± 0.4
T-P (mg/L) n.a. 0.055 0.064 ± 0.01
*Not available
The runoff water from the conventional and SRI water management plots showed different
NPS pollutant concentrations and loads as seen in Table 3. The concentration and load of SRI-30 and
SRI-40 were not significantly different, and they were accordingly pooled and averaged in Table 3.
The average concentrations and loads of the CT and CD plots were significantly higher than those of
the CS and SRI plots at the 5% level. Reviewing the different measured water quality indices, we see
NPS pollution load reductions by SRI water management ranging from 16.5 to 53.9% in the CS plots,
and from 27.1 to 46.0% in the SRI plots. Total nitrogen (TN) showed the lowest reduction. A large
amount of soluble nitrate nitrogen (NO3-N) might have been easily discharged with the draining water,
thus resulting in the low TN reduction. The differences in the concentrations and loads between the CS
and SRI plots were not significant. The results suggest that SRI water management in Korean paddy
fields could significantly decrease both the concentration and load of NPS pollution and contribute to
improving water quality in rivers and lakes.
Table 3 Comparison of average NPS pollution concentration and load with respect to experimental
treatments in the 2011 growing season
Water quality index BOD SS CODMn CODCr T-N T-P
CT 3.6a 59.2a 8.2a 22.0a 5.09a 0.45a
CD 3.3a 46.6a 7.5a 20.5a 5.20a 0.48a
Concen-
CS 2.9b 32.8b 5.6b 15.3b 4.85ab 0.32b
tration
SRI 2.55b 33.80b 5.75b 16.15b 4.13b 0.34b
(mg/l)
Reduction CS:CT 19.4% 44.6% 31.7% 30.5% 4.7% 28.9%
Reduction SRI:CT 29.2% 42.9% 29.9% 26.6% 19.0% 25.6%
CT 56.8a 1,031.4a 119.8a 324.7a 77.0a 5.8a
CD 43.8a 1,036.3a 106.1a 327.3a 81.6a 6.1a
Load CS 34.8b 562.6b 55.2b 168.0b 64.3b 3.4b

(kg/ha) SRI 30.6b 583.4b 70.1b 210.8b 56.1b 3.5b
Reduction CS:CT 38.7% 45.5% 53.9% 48.3% 16.5% 41.4%
Reduction SRI:CT 46.0% 43.4% 41.5% 35.1% 27.1% 39.7%
*Values having different letters in column are significantly different at 5% significance level
3.2 Greenhouse gas emissions

Fig. 2 and 3 show the temporal methane (CH4) and nitrous oxide (N2O) emissions from the
paddy plots in terms of experimental treatments and water temperature. The methane emissions from
the CT plots were higher than those from the CD and CS plots throughout the growing season. The
methane emissions from the CS plots were higher at the beginning of rice cultivation before they
decreased and the lower emissions rates were then maintained. Further, the methane emissions from
the CD plots seemed to be affected by the two forced drainages. The nitrous oxide emissions showed
different emission patterns from those for methane. Low nitrous oxide emissions were measured from
the CT plots throughout the growing season. As in the CT plot, nitrous oxide emissions from the CD
and CS plots were higher in the first half of the growing season before they decreased and lower
emissions were maintained.
Fig. 2 Temporal methane (CH4) emission at the paddy plots with respect to experimental treatments and
water temperature
Fig. 3 Temporal nitrous oxide (N2O) emissions at the paddy plots with respect to experimental treatments
and water temperature
The methane and nitrous oxide emissions can be converted to carbon dioxide (CO2) equivalents
by multiplying them by 21 and 310, respectively, which reflect the respective global warming potential
(GWP) of these molecules (IPCC, 1996). Table 4 shows the greenhouse gas emission with respect to
the CT, CD and CS treatments. The methane emissions from the CD and CS plots were decreased
significantly, to 157.7 kg/ha and 126.8 kg/ha, respectively, compared with 458.4 kg/ha from the CT
plots. At the same time, the increase in N2O emissions from the CD and CS plots compared with those
from the CT plots was relatively small and did not offset the decrease in methane emissions if
converted to CO2 equivalent emissions. As such, the GHG emissions from the CD and CS plots in
terms of their CO2 equivalents were reduced by 65.5% and 72.8%, respectively, compared with those
from the CT plots.
Table 4 Comparison of CH4 and N2O emission between different water management
Emissions (kg/ha) CO2 ton/ha Index
Treatment
CH4 N2O equivalent (%)
CT 458.4 0.000028 14.2 100
CD 157.7 0.007 4.9 34.5
CS 126.8 0.074 4.0 28.2
Methane-producing bacteria are usually very sensitive to the presence of oxygen in the soil. As
such, when the soil becomes aerated, these bacteria died off and could not recover their population
quickly, resulting in large reductions in methane emissions. The reduction measured was larger than
that indicated by a World Bank (2009) report which claimed that 30% less methane would be emitted
from rice paddies if they are drained. This report contradicted a previous report by Kasimir-
Klemedtsson et al. (1997), which had claimed that if peat soil wetlands are drained for agricultural
production, the resulting increase in CO2 and N2O emissions would far exceed the benefits from
suppression of methane emissions. This is a complicated matter as greenhouse gas emissions vary
according to factors such as soil texture, organic matter and carbon content, water management, and
other practices in paddy farming. So multiple studies need to be conducted to assess the net effects on
GHG emissions under various soil and cropping conditions.
However, our trials indicate that in Korean rice farming, if the paddy soils are aerated by either
SRI water management or forced drainage, the benefits of methane emission reductions will exceed
the disadvantages of increased nitrous oxide emissions. An application of SRI water management in
Korea appears to reduce GHG emissions significantly, thereby contributing to achieving the
government's goal of GHG reduction in the agricultural sector to mitigate global warming.
3.3 Crop performance effects

The number of tillers per hill in the SRI and CS plots was larger than that in the CT and CD
plots (Table 5). Further, the number of tillers in the SRI and CS plots increased as the transplant
spacing and the number of seedlings per hill increased. The water management with two forced
drainages (CD) did not seem to contribute to an increase in the number of tillers per hill. As seen in the
CS plots, the SRI water management itself also did not seem to increase the number of tillers per hill.
Narrow transplant spacing might have been a main cause for the low tillering. It has been thought that
the number of tillers per hill for Japonica rice under SRI cultivation is generally lower than that of
Indica rice. The largest numbers of tillers per hill in this study were 53 and 60, respectively, with three
seedlings transplanted per hill and spaced 30 cm and 40 cm, respectively. When one seedling per hill
was transplanted, the numbers of tillers per hill were only 38 and 41, respectively. Based on the
observation for larger tillering, two to three seedlings per hill transplanting were thought to be
advantageous if a Japonica rice variety is cultivated under the SRI method. The plant heights in the CS
and SRI plots were taller than that in the CT plots. However, the difference in height between the CS
and SRI plots was not significantly different at the 5% level.
Table 5 Number of tillers per hill and plant height with respect to treatment and measured date
Measured date (2011)
Treatment
27 May 1 Jun 8 Jun 22 Jun 1 Jul 20 Jul 5 Aug
CT 7 8 12 21 30 33 33
CD 7 10 13 22 30 34 32
CS 6 8 12 21 33 37 35
SRI-30/1 7 9 11 23 32 41 38
No. of
SRI-30/2 13 16 18 35 47 52 48
tillers/hill
SRI-30/3 19 24 30 41 52 54 53
SRI-40/1 7 10 12 24 33 42 41
SRI-40/2 13 17 22 36 52 54 52
SRI-40/3 19 25 32 45 63 68 60
CT 16 20 26 41 50 85 104
CD 19 23 28 44 52 82 106
CS 16 20 27 47 62 86 108
Plant SRI-30/1 18 24 31 48 63 86 106
height SRI-30/2 20 23 31 49 63 89 110

(cm) SRI-30/3 18 24 33 50 67 91 108
SRI-40/1 20 23 31 50 61 88 108
SRI-40/2 18 23 33 50 64 92 112
SRI-40/3 20 25 33 52 68 92 110
The yields of polished rice from all the experimental plots in this study were generally higher
than the common yields of 4.50~4.70 ton/ha from the cultivar Odaebyeo in Korea. Application of
vermiculture compost to all of the plots in addition to the recommended chemical fertilizer application
was thought to be one of reasons for the higher yields across the board. The polished rice yields in the
SRI and CS plots were significantly higher than those from the CT plots (Table 6). The two highest
polished rice yields of 6.47 ton/ha and 6.34 ton/ha were observed in the SRI-30-3 and CS plots,
respectively, representing respective increases of 20% and 17% compared with those in the CT plots.
If head grain yields (unbroken grains coming from milling) for SRI-30-3 and CS are compared with
CT, the increases were 38% and 28%, respectively. This indicates that the rice in the SRI water
management plots matured better than in the CT plots and had fewer broken grains resulting from the
polishing processes.
Table 6 Comparison of rice yield properties with respect to different treatments
Effective Weight/ Yield (kg/10a) Yield index
Weight/ Panicles/ Grains/
Treatment grains/panicle 1000 Polished Head Polish Head
hill (g) hill panicle
(%) grains(g) rice rice ed rice rice
CT 34.7 21.5 71.5 86.3 25.9 541.0 431.0 100.0 100.0
CD 35.4 22.0 69.0 85.8 26.1 562.0 465.0 104.0 108.0
CS 41.4 27.0 80.0 90.2 26.1 634.0 552.0 117.0 128.0
SRI-30-1 71.7 31.5 95.5 94.5 27.2 611.0 532.0 113.0 124.0
SRI-30-2 73.0 34.0 93.5 94.2 26.8 612.0 533.0 113.0 124.0
SRI-30-3 74.9 38.0 84.0 93.1 25.0 647.0 574.0 120.0 133.0
SRI-40-1 123.2 44.5 106.0 91.6 27.7 590.0 518.0 109.0 120.0
SRI-40-2 125.2 51.0 96.0 92.5 26.9 591.0 505.0 109.0 117.0
SRI-40-3 128.9 55.0 92.0 93.1 26.4 627.0 542.0 116.0 130.0
The rice weight per hill increased as the number of seedlings per hill and transplant spacing
increased. The planting density of the SRI plots was much lower than that of the CS plots overall.
Therefore, it was thought that the highest possible rice yield could be achieved if the planting density
and transplant spacing were to be better optimized. The usual recommendation for SRI practice is
25x25 cm spacing of single seedlings. The highest yield was observed at the SRI plots, with transplant
spacing of 30 cm and three seedlings planted per hill. However, in practical terms, transplanting three
seedlings per hill at a spacing of 7 cm in a triangular shape is practically impossible in Korea’s current
rice culture, even though it is now the preferred spacing for SRI in Sichuan province of China, where
SRI use has increased from 1,133 ha in 2004 to over 300,000 ha in 2010 (Zheng et al., 2011). Further
studies on the effect of transplant spacing and number of seedlings per hill are recommended to find an
optimizing combination between the two. The current conventional transplanting spacing of 30x15 cm
with four to six seedlings per hill under SRI water management was thought to be a good practice for
both better rice yields and reductions of NPS pollution and GHG emissions. This practice could be
easily adopted by Korean rice farmers without changing and modifying conventional rice farming
methods and their existing farm machinery.
4. Conclusions
An experimental field plot study was performed in 2011 to measure the effect of paddy
irrigation management in Korea on rice yields, water use, NPS pollution discharges, and GHG
emissions. Twelve plots sized 5x15 m were prepared, and a locally bred Japonica rice cultivar was
cultivated. Experimental treatments were conventional cultivation (CT), conventional cultivation with
SRI water management (CS), conventional cultivation with two forced mid-season drainages (CD),
and the SRI method with transplant spacing between hills of 30x30 cm (SRI-30) and 40x40 cm (SRI-
40). Each treatment was replicated. The rice yields, water use, water quality (SS, BOD5, CODCr,
CODMn, TN and TP) of the irrigation and drainage water, GHG (CH4 and N2O) emissions, and other
necessary variables were measured and analyzed according to the relevant standards and procedures.
The soil texture of the plots was a loam. The organic matter content was 28±1 g/kg. The
measured average infiltration and evapotranspiration of the plots were0.3 mm/day and 4.6 mm/day,
respectively, indicating that evapotranspiration was a major mechanism consuming water in the paddy
plots and that the infiltration loss was insignificant. During the growing season of May to September, a
total rainfall of 1,609.9 mm was recorded at the plots.
A very high rate of irrigation reduction was achieved using the SRI and CS treatments at 49.4%
and 47.6%, respectively. The quality of the irrigation water was within the national standard and did
not show large differences from the national average in general. The NPS pollution load reductions by
SRI water management ranged between 16.5% and 53.9% from the CS plots and 27.1% and 46.0%
from the SRI plots depending on measured water quality indices. The total nitrogen (TN) showed the
lowest reduction.
The methane emissions from the CD and CS plots decreased significantly to 157.7 kg/ha and
126.8 kg/ha, respectively, compared with 458.4 kg/ha from the CT plots. However, the increase of
N2O emissions from the CD and CS plots compared with that from the CT plots was not comparable
with the methane emission decrease. The GHG emissions from the CD and CS plots in terms of CO2
equivalence, were reduced by 65.5% and 72.8%, respectively, compared with that from the CT plots.
The numbers of tillers per hill in the SRI and CS plots were larger than those in the CT and CD
plots, and increased as transplant spacing and number of seedlings per hill increased. The largest
number of tillers per hill in the SRI plots was 53 and 60, respectively, where three seedlings per hill
(spaced 30 cm and 40 cm) were transplanted. The polished rice yields in the SRI and CS plots were
significantly higher than those in the CT plots. The two highest polished rice yields were 6.47 ton/ha
and 6.34 ton/ha from the SRI-30-3 and CS plots, respectively, representing respective increases of 20%
and 17% compared with that from the CT plots.
It was concluded that SRI water management in Korean paddy farming could significantly
decrease the irrigation water requirements, NPS pollution discharges and GHG emissions and thus
help save irrigation water, improve the national irrigation system, contribute to improving water
quality in rivers and lakes and mitigate global warming and climate change. Further studies on the
effects of transplant spacing and number of seedlings per hill may find the best combination of the two
for higher rice yields. The CS treatment could be easily adopted by rice farmers, who would not have
to change or modify their conventional rice farming methods and existing farm machinery to partly
achieve the effects of SRI.
Acknowledgements
This research was supported by Rural Research Institute of Korea Rural Development
Corporation, the Ministry for Food, Agriculture, Forestry and Fisheries, Korea. The authors appreciate
their support.
References
Bonta, J. V, 2002, “Modification and performance of the Coshocton wheel with the modified drop-box
weir.” Journal of Soil Water Conservation, 57(6), 364-373.
Bonta, J. V., and Pierson, F. B., 2003, “Design, measurement, and sampling with drop-box weirs.”
Applied Engineering in Agriculture, 19(6), 689-700.
Chapagain, T., and Yamaji, E., 2010, “The effects of irrigation method, age of seedling and spacing on
crop performance, productivity and water-wise rice production in Japan.” Journal of the
International Society of Paddy and Water Environment Engineering, 8, 81-90.
Choi, J. D., Park, W. J., Park, K. W., and Lim, K. J., 2013, “Feasibility of SRI methods for reduction of
irrigation and NPS pollution in Korea.” Journal of the International Society of Paddy and Water
Environment Engineering, 11(1-4), 241-248.
Dixit, K., 2005, The miracle is it’s no miracle. The Nepali Times. Kathmandu.
http://www.nepalitimes.com/ issue/256/Nation /569.
FAO.,2011, Climate change, water and food security. FAO Water Reports, 36, 119-128.
Freibauer, A., Rounsevell, M. D. A., Smith, P., and Verhagen, J., 2004, “Carbon sequestration in the
agricultural soils of Europe.” Journal of Geoderma, 122(1), 1-23.
Hameed, K. A., Mosa, A. K. J., and Jaber, F. A., 2011, “Irrigation water reduction using System of
Rice Intensification compared with conventional cultivation methods in Iraq.” Journal of the
International Society of Paddy and Water Environment Engineering, 9(1), 121-127.
IPCC., 1996, Revised IPCC guideline for national greenhouse gas inventories: Reference manual
Karki, S., 2010, System of rice intensification: An analysis of adoption and potential environmental
benefits. Master's thesis. Norwegian University of Life Sciences, Ås, Norway. August, 2010
Kasimir-Klemedtsson, A., Klemedtsson, L., Berglund, K., Martikainen, P., Silvola, J., and Oenema,
O. , 1997, “Greenhouse gas emissions from farmed organic soils: a review.” Journal of Soil Use
and Management , 13(s4), 245-250.
Krupnik, T. J., Shennan, C., and Rodenburg, J., 2012, Yield, water productivity and nutrient balances
under the System of Rice Intensification and Recommended Management Practices in the Sahel.
Journal of Field Crops Research, 130, 155-167.
Ministry of Environment., 2007, Standard Methods for Examination of Water Quality, Gwacheon,
Korea
Ministry of Environment, 2009, Standard Methods for Examination of Soil Pollution, Gwacheon,
Korea
Noborio, K., Kudo, Y., Kato, T., and Shimoozono, N., 2012, “Effect of water management practices on
paddy rice yield and emission of greenhouse gases.” Presentation at the 2012 CIGR Congress,
Paper No. P-1892. July 8-11, 2012. Valencia, Spain.
Nyamai, M., Mati, B. M., Home, P. G., Odongo, B., Wanjogu, R., and Thuranira, E. G., 2012,
“Improving land and water productivity in basin rice cultivation in Kenya through System of Rice
Intensification.” Journal of CIGR, 14(2), 1-9.
Thakur, A. K., Rath, S., Roychowdhury, S., and Uphoff, N., 2010, “Comparative performance of rice
with System of Rice Intensification (SRI) and conventional management using different plant
spacings.” Journal of Agronomy and Crop Science, 196(2), 146-159.
Thakur, A. K., Rath, S., Patil, D. U., and Kumar, A., 2011, “Effects on rice plant morphology and
physiology of water and associated management practices of the system of rice intensification
and their implications for crop performance.” Journal of the International Society of Paddy and
Water Environment Engineering, 9(1), 13-24.
Uphoff, N., Kassam, A., and Harwood, R., 2011, “SRI as a methodology for raising crop and water
productivity: Productive adaptations in rice agronomy and irrigation water management.” Journal
of the International Society of Paddy and Water Environment Engineering, 9(1), 3-11.
World Bank., 2009, World Development Report 2010: Development and climate change. World Bank,
Washington D.C., USA.
Yagi, K., 1991, “Emission of biogenic gas compounds from soil ecosystem and effect of global
environment. 2. Methane emission from paddy fields.” Journal of Soil and Fert., 62(5), 556-562.
Zheng, J.G., Lu, X., Jiang, X., and Tang, Y., 2004, The system of rice intensification (SRI) for super-
high yields of rice in Sichuan Basin. International Agronomy Conference. Paper presented at the
4th International Crop Science Congress in Brisbane, Australia. http://www.cropscience.org.au/
Zheng, J.G., Zhou, L., Chi, Z.Z., and Jiang, X.U., 2011, Agricultural Water Savings Possible through
SRI for Future Water Management in Sichuan, China. http://sri.ciifad.cornell.edu/
THE FUTURE
E-04
The Impact of Agriculture Policy to Rural Water

Management in Northern Taiwan
Ray-Shyan Wu, Chia-Chi Ma

* Dept. of Civil Engineering, National Central University, Taoyuan, Taiwan
ABSTRACT
During earlier periods, the social economy of Taiwan is predominantly based on agriculture. But with
the economic development, industrial structure changing, the GDP of agricultural output was declined
from 32% in 1951 to 2% in 2011. The plantation areas of rice, the mainly food crop of Taiwan, was
79*104 ha in 1975. Since then, the plantation areas were declined to 25*104 ha in 2011. Agricultural
policy would guide the farmers’ production willingness and agricultural land use situation directly, that
is, impacting the agricultural water management directly. In recent years, the greatest impact on
agricultural water management in agricultural policy was “The program of water-resources
distribution concerning the adjustments of rice-paddy and upland crop fields” in 1997. This program
was aimed at adjusting the structure of rice production and marketing in order to join the World Trade
Organization. The plantation areas of rice were declined from 36.4*104 ha in 1997 to 26.9*104 ha in
2005. Another policy greatly influenced for water management was “Farming practices adjustment and
farmland renovation program” in 2013. In order to reduce the international food supply risks caused by
climate change, this policy would guide the farmers in 5*104 ha fields who lying fallow for years to
planting several grains. It was expected to improve the domestic food self-sufficiency rate from 33.5% in
2012 to 34.9% in 2016 and maintaining the agricultural production environment. In this study, we take
advantage of the water balance model to investigate the above two major agricultural policies on the
impact of agricultural water management. In addition to meet irrigation needs, but also to be discussed
on regional water resources regulation and the potential impact of drought resistance and flood
detention.
Keyword agricultural policy, agriculture water management
1. Introduction
During earlier periods, the social economy of Taiwan are predominantly based on agriculture.
But with the economic development, industrial structure changing, the GDP of agricultural output was
declined from 32% in 1951 to 2% in 2011. The plantation areas of rice, the mainly food crop of
Taiwan, was reached a peak of 79*104 ha in 1975. Since then, the plantation areas was declined with
the changes in the industrial structure gradually reduced. The government adjusted the rice production
and marketing structure in order to join the World Trade Organization. The Council of Agriculture
started “The program of water-resources distribution concerning the adjustments of rice-paddy and
upland crop fields” in 1997. The plantation areas of rice was declined from 36.4*104 ha in 1997 to
25.4*104 ha in 2011. However, we compared the agricultural water use over the years, it wasn’t with a
proportional decrease in irrigated area. In the meantime, domestic and industrial water demand is
increasing year by year. The agricultural water were required to reduce because of new water sources
developing were difficult and higher costs in natural or social conditions.
The Council of Agriculture set the policy objectives for food self-sufficiency rate of 40 % before
2020 in the "National Food Security Meeting" in 2011. For ensuring the food security, they will
adjust the domestic agriculture production structure, and improve the using efficiency of agricultural
land and water resources. The medium-term project of adjusting the cropping system and rearranging
the farmlands started in 2013. It coach the continuous fallow land landowners to rehabilitate for one
crop-term or rent to others to crop industrial crops. It expected to improve the food self-sufficiency
rate of 34.9% in 2016, and improve the total agricultural output value of 8.8 billion dollars, and also
reduce the carbon dioxide emissions by 61,772 tons.
Because of the agriculture policy will guild the willingness of production for farmers and using
situation for farms directly, it will impact on agricultural water management directly. “The program of
water-resources distribution concerning the adjustments of rice-paddy and upland crop fields” and “The
medium-term plan of adjusting the cropping system and rearranging the farmlands” are the greatest
impacting agricultural policy on agricultural water management in recent years. In this paper, Taoyuan
Irrigation Association for the study area, explored cultivated land area and water management changes
under the guidance of the agricultural policy.
2. Research Area
The irrigation area of Taoyuan Irrigation Association (TIA) is located in the northern of Taiwan,
which includes Taipei, Taoyuan, and Hsinchu counties. It can be divided into four irrigation areas,
namely Taoyuan, Hukou, Tahsi and Hsinhai according to the physical features and irrigation system.
The total irrigation area is 24,650 hectares. The irrigation water resource was Shihmen Reservoir,
several rivers and effective rainfall. The climate in this region is sub-tropical climate with humid and hot
weather. Therefore, crops grow all the year long. The annual average temperature is 21℃. The summer
is long and a high temperature with an average of 27.6℃. The winter is short and lower temperature with
an average of 15℃. The annual average rainfall is 2,000mm. uneven distribution of rainfall in different
seasons that it has more rainfall in summer than in winter. The effective rainfall in the irrigation season
makes up 15% of the total amount of water irrigated. The main crop occupying the 90% of farmland in
the irrigation area is paddy rice, which can be grown twice a year. The remained 10% of farmland in this
area is planted with sweet potatoes, melons, vegetables, etc.
The irrigation area of TIA was divided into 329 irrigation groups. The irrigation water resources
were reservoir supply of 56%, rivers supply of 29%, and effective rainfall of 15%. The TIA
administrated 284 ponds, the unique irrigation facilities in Taiwan, to regulate the irrigation water. The
total water storage was 4.6 * 107 m3. The TIA regulated irrigation water in the ponds cooperated with
Shihmen Reservoir and rivers. Stored the irrigation water in the ponds first, and irrigated farms for the
right time and right amount. It would improve the efficiency of rainfall, and irrigate stably.
Taipei City
Taoyuan County
Taoyuan Canal
Shihmen Reservoir
Fig. 1 Irrigation area aerial photograph of Taoyuan Irrigation Association
3. The Program Of Water-Resources Distribution Concerning The Adjustments Of

Rice-Paddy And Upland Crop Fields
The Program was Taiwan’s response to the accession to the World Trade Organization, for
adjusting the industrial structure of rice, cereals and deed sugar cane. It started in 1997. The plantation
area of rice was declined from 36.4*104 ha to 30.7*104 ha in 2002. The plantation area of cereals was
declined from 4.5*104 ha to 2.1*104 ha. The plantation area of deed sugar cane was declined from
2.2*104 ha to 1.2*104 ha.
Taiwan joined the World Trade Organization in 2002, and allowed to import some rice. Therefor,
the area of fallow was increased year by year. The plantation area of rice was declined to 26*104 ha in
2012. But this program focused on rice area and yield adjustments without centralized fallow area,
didn’t reduce the irrigation water in the meantime. The farmers cropped the rice in the dry 1st crop-term
and rested in the wet 2nd crop-term for the higher production value of 1st crop-term rice, it reduced the
efficiency of water resources.
In the Taoyuan area, the plantation area of rice was 3.97*104 ha in 1997, declined to 2.09*104 ha
for joined the WTO in 2002. And declined further to 1.4*104 ha in 2007. Next few years remained at
1.2*104 to 1.4*104 ha.
Table 1 The Execution results of “The program of water-resources distribution concerning the adjustments
of rice-paddy and upland crop fields”
Maintain the The plantation areas of rice was declined from 36.4*104 ha in 1997 to 26.9*104 ha
domestic balance of in 2005, and 25.4*104 ha in 2011. The policy objectives of maintaining domestic
rice supply and balance of rice supply and demand was achieved.
demand
The annual average price per kilogram of Japonica Rice was $18.68 for the
Stabilize the market previous 3 years by 2001, and increased to $22.12 in 2009. It was reached the most
prices ideal state for adjusting the industrial structure of rice and stabilized the market
prices effectively.
The average farmer earning of 1st crop rice was $55,772 for the previous five years
Ensure the income
by 2001, and increased to $60,407 for the last five years. It was ensured the farmers
of farmers
income effectively.
Reduced the AMS of the rice of public grain from 4.82 billion in 2001 to 2.77
Reduce the domestic
billion in 2008. We compared the AMS in 2008 to the base year of 2001, the
support for
domestic AMS has dropped from 17.71 billion to 4 billion, a drop of 77%, reaching
agriculture (AMS)
reduction goal.
Table 2 The Comparison of rice supply and demand of Taiwan for joining the WTO
Area of Rice Population Total Consumption of Total Supply of Taiwan
year of Taiwan of Taiwan Taiwan (103 ton)
(103 ha) (103 ) (103 ton) importation production
1991 429 20,286 1,519 0 1,819
2001 332 22,278 1,269 0 1,396
2002 307 22,396 1,273 144 1,461
2006 263 22,823 1,245 144 1,261
2009 255 23,042 1,243 144 1,276
2011 254 23,225 1,188 144 1,348
2012 260 23,316 144 1,368
Table 3 The plantation area and irrigation water of rice of TIA in 1995~2012
1st crop-term 2nd crop-term
Area
Reservo River Field Area of Reservo River Field
year of Irrigatio Irrigatio Irrigatio Irrigatio
ir supply irrigatio rice ir supply irrigatio
rice n depth n loss n depth n loss
supply (106 3 n depth (103 supply (106 n depth
(103 (mm) (10 6
m ) (mm) (106 m3)
(106 m3) m3) (mm) ha) (106 m3) m3) (mm)
ha)
1995 22.5 179 91 1,200 90 800 21.9 157 124 1,283 97 842
*
1996 13.8 31 80 804 41 505 21.3 159 103 1,230 89 813
1997 22.2 180 87 1,203 89 803 20.7 175 109 1,372 96 908
1998 21.1 184 83 1,265 88 846 20.7 175 108 1,367 96 905
1999 21.5 183 107 1,349 98 894 21.1 205 138 1,626 117 1073
2000 22.3 184 102 1,283 96 852 20.9 171 107 1,330 94 880
2001 21.7 198 93 1,341 97 896 20.5 137 182 1,556 114 1000
*
2002 19.8 101 122 1,126 79 727 17.2 143 120 1,529 91 1001
*
2003 1 0 0 0 0 0 6 94 85 2,983 62 1947
*
2004 0.9 0 0 0 0 0 6.1 121 44 2,705 54 1821
2005 12.4 143 54 1,589 65 1069 6.5 123 52 2,692 58 1805
*
2006 3.7 28 15 1,162 14 773 5.1 106 86 3,765 66 2467
2007 9.2 129 76 2,228 69 1477 5.3 137 59 3,698 65 2477
2008 9.3 130 70 2,151 67 1430 4.9 131 59 3,878 63 2594
**
2009 6.8 116 62 2,618 60 1741 4.3 126 49 4,070 57 2735
2010 9.2 135 58 2,098 64 1405 6.1 143 49 3,148 63 2123
**
2011 6.8 118 70 2,765 63 1832 6 155 59 3,567 70 2398
2012 7.4 157 84 3,257 81 2166 6 154 68 3,700 73 2477
* : It was dry during the 1st crop, the irrigation water and rice cropping was limited by the government.
** : It was dry during the 1st crop, only the irrigation water was limited by the government.
200 100
Area (103 ha)
Water Supply (106 m3)
150 80
60
100
40
50 20 Reservoir
Supply
0 0
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
year
Fig. 2 Irrigation Water Use and Rice Area of TIA
Excluding the dry year, and to join the WTO in 2002 for the sector, the average rice area for 1st
crop-term in 1995~2001 was 21.9*103 ha, the average reservoir supply was 185*106 m3, meaning the
average field irrigation depth was 848mm. The average rice area for 1st crop-term in 2002~2012 was
9.5*103 ha, the average reservoir supply was 139*106 m3, meaning the average field irrigation depth was
1,455mm. After joining the WTO , the rice area had dropped to 43% of the original, and the reservoir
supply and river supply were reduced to 75% and 73% in the meantime.
Comparing the field irrigation depth with the irrigation plan of TIA of 1,070mm, the field irrigation
depth before joining the WTO was 848mm for the more rice area. To meet the demand for irrigation, the
TIA need to perform more sophisticated rotation irrigation , as well as strict control of water distribution
and improve the effective utilization of rainfall. After joining the WTO, the rice area had reduced
significantly, and the field irrigation depth increased to 1,455mm, meaning the irrigation water was
ample.
The TIA has the unique irrigation facilities of 284 ponds, and the total water storage was 4.6 * 107 m3. It
would coordinate with Shihmen Reservoir to store and regulate the irrigation water. In particular in the
dry 1st crop-term, it could improve the efficiency of spring rainfall and delay or reduce the irrigation
water supply for Shimen Reservoir to save and store the water resources. Though the rice area had
reduced significantly under the agriculture policy guiding, the decreasing ratio of the irrigation water
was less than the rice area. But the abundant irrigation water would contribute to the environment during
the wet season and lurk greater support capability for the livelihood and industrial water during the dry
season.
For the examples of dry 1st crop-term in 2009 and 2011, the rainfall in Taoyuan area were only 57% and
79% of the average of the last decade. The Water Resources Agency controlled the water allocation of
Shimen Reservoir from the beginning of the 1st crop-term. As the reservoir supply for irrigation was
declined to 67% of the planning value, the TIA were required to support the Water Company for the
livelihood and industrial water. The support amount were 22*106 m3 in 2009 and 17*106 m3 in 2011. As
the saving of irrigation water were 54*106 m3 in 2009 and 53*106 m3 in 2011, the real utilization of
irrigation water were 53% in 2009 and 57% in 2011.
Table 4 Water Conservation of TIA in 2009 and 2011

Planned Real
Reservoir Support for Real irrigation
irrigation utilization of Water Saving
year supply Water Company water
water demand irrigation (106 m3)
6 3 (106 m3) (106 m3) (106 m3)
(10 m ) water
2009 163 109 22 87 53% 54
2011 163 110 17 93 57% 53
4. The Medium-Term Project of Adjusting the Cropping System and Rearranging the
Farmlands
The Council of Agriculture set the policy objectives for food self-sufficiency rate of 40 % before
2020 in the "National Food Security Meeting" in 2011. For ensuring the food security, they will adjust
the domestic agriculture production structure, and improve the using efficiency of agricultural land and
water resources. The policy goals included mastering the international sources of food imports to
strengthen the international agricultural investment and cooperation, building the food security
classification management system, and mastering the food security stocks.
In order to reduce the international food supply risks caused by climate change, this agriculture
policy would guide the farmers whose fields lying fallow for years to rehabilitate for one crop-term or
rent to others to crop industrial crops. The recommended crops included the organic crops, regional
specialties, import-substitution crops, and crops with export potential such as feed corn, soybeans, peas,
lettuce, carrots, and short-term economic forest. It was expected to improve the domestic food
self-sufficiency rate of 34.9% in 2016 and improve the total agricultural output value of 8.8 billion
dollars, and also reduce the carbon dioxide emissions by 61,772 tons.
This project’s primary objectives are excellent farms lying fallow for all crop-terms by years. The
recommended crops were drought-resistant crops whose irrigation demand less than the rice. We
estimated the annual irrigation water demand for rehabilitation was 6.46*108 m3. Comparing with the
irrigation water supply last decade, the irrigation water demand will increase 16.8% in southern of
Taiwan as 6% in northern of Taiwan. Although the irrigation water rights able to support the current
irrigation water supply plus the water demand for rehabilitation, but the drought-support capability will
decline relatively. There are heavier burden for the management of water resources, in particular on the
dry 1st crop-term.
The rehabilitated farm area was 12*104 ha in 2013 of the first year of the project. The increasing
area of rice was 9.8*104 ha, the area of recommended crops just only 2.2*104 ha. Moreover the main
increased area was located on the area with higher risk of water shortage such as Taoyuan, Yunlin and
Tainan. Therefore the rice area for 1st crop-term was 12,275 ha in Taoyuan area, 132% for the last
decade. We could classify the rice location according to the irrigation water source such as reservoir
supply, rivers supply, and hybrid supply. The rice area and field irrigation depth of the TIA showed in
Table 5.
Table 5. The Rice Area and Field Irrigation Depth of TIA in 2013 and the Last Decade
Rice Area & FID by
Rice Area & FID by Rice Area & FID by Total Rice Area &
Reservoir & River
Reservoir Supply River Supply FID
Supply
Average 1st
5,669 ha 3,356 ha 470 ha 9,495 ha
Crop-term in
1,371 mm 982 mm 5,847 mm 1,455 mm
2002~2012
1st Crop-term in 7,898 ha 4,037 ha 592 ha 12,527 ha
2013 893 mm 795 mm 4,957 mm 1,054 mm
139% 120% 126% 132%

Ratio
65% 81% 85% 72%
FID: field irrigation depth
Due to the spring without rains, the rainfall within January to March on the watershed of Shihmen
Reservoir was only 18% of the historical averages. The water elevation of the reservoir has been
dropped from 244.6m as full elevation of January 1st to 233m of March 12th. The effective storage of the
reservoir has been dropped from 99% to 52%. The drought situation was more serious than the 2009 and
2011. Therefore the Water Resources Agency controlled the water allocation of Shimen Reservoir from
March 1st, the reservoir supply for irrigation was declined to 75% of the planning value. Because of the
significant increasing in rice area, the TIA encounter several difficulties on irrigation water allocation.
Fortunately, it rained as scheduled in early May, the reservoir supply satisfied the irrigation water as
irrigation plan. The irrigation water supply for the whole 1st crop-term still satisfied 96% averages of last
decade. While the field irrigation depth didn’t meet the plan depth due to the rice area increasing, the
irrigation association will accomplish the irrigation plan with their experience of water management.
5. Discussion
In recent years in Taiwan, the greatest impact on agricultural water management in agricultural
policy was “The program of water-resources distribution concerning the adjustments of rice-paddy and
upland crop fields” in 1997. This program was aimed at adjusting the structure of rice production and
marketing in order to join the World Trade Organization. The plantation areas of rice was declined from
36.4*104 ha in 1997 to 26.9*104 ha in 2005. For the research area, excluding the dry year and joining the
WTO in 2002 for the sector, the average rice area for 1st crop-term in 1995~2001 was 21.9*103 ha, the
average reservoir supply was 185*106 m3, meaning the average field irrigation depth was 848mm. The
average rice area for 1st crop-term in 2002~2012 was 9.5*103 ha, the average reservoir supply was
139*106 m3, meaning the average field irrigation depth was 1,455mm. After joining the WTO , the rice
area had dropped to 43% of the original, and the reservoir supply and river supply were reduced to 75%
and 73% in the meantime.
Though the rice area had reduced significantly under the agriculture policy guiding, the decreasing
ratio of the irrigation water was less than the rice area. But the abundant irrigation water would
contribute to the environment during the wet season and lurk greater support capability for the
livelihood and industrial water during the dry season. For the examples of dry 1st crop-term in 2009 and
2011, the reservoir supply for irrigation was declined to 67% of the planning value, the TIA were
required to support the Water Company for the livelihood and industrial water. The support amount
were 22*106 m3 in 2009 and 17*106 m3 in 2011. As the saving of irrigation water were 54*106 m3 in
2009 and 53*106 m3 in 2011, the real utilization of irrigation water were 53% in 2009 and 57% in 2011.
In order to reduce the international food supply risks caused by climate change and ensure the food
security, the Council of Agriculture will adjust the domestic agriculture production structure, and
improve the using efficiency of agricultural land and water resources. The medium-term project of
adjusting the cropping system and rearranging the farmlands started in 2013. The rehabilitated farm area
was 12*104 ha in 2013 of the first year of the project. The increasing area of rice was 9.8*104 ha, the area
of recommended crops just only 2.2*104 ha. Moreover the main increased area was located on the area
with higher risk of water shortage such as Taoyuan, Yunlin and Tainan. The rice area for 1st crop-term
was 12,275 ha in Taoyuan area, 132% for the last decade. Therefore the field irrigation depth declined to
72%, in particular in the area of reservoir supply of 65%.
6. Conclusion
Targeting on the water-resources distribution concerning the adjustments of rice-paddy and upland
crop fields following the accession to the WTO, the Water Resources Agency has been promoting
related studies as well as projects, and has achieved preliminary results. However, for the
comprehensive consolidation of the result, it is necessary to proceed with in-depth studies, in order for
the successful distribution of agricultural water as an aid to the reliable public water-supply. Current
implementation of the project as well as fallow area was exercised by following the administrative
system, that is, townships were responsible for execution under the supervision of county governments.
However, when the goal on the saving of agricultural water is added, it is suggested to follow the local
irrigation systems, that is, work stations or lateral would be responsible for the execution under the
negotiation of irrigation associations. In other words, current regulations or procedures have to be
reviewed and modified.
“The Medium-Term Project of Adjusting the Cropping System and Rearranging the Farmlands”
recommended the farmers to plant the less water-required crops such as organic crops, regional
specialties, import-substitution crops, and crops with export potential for the original ideas. And
rehabilitated the fallows in the wet 2nd crop-term. But the actual rehabilitated fallows were rice paddy of
83%, and cropped in the dry 1st crop-term of 53%. The risk of water management for the dry season had
increased. Due to the agricultural policy have considerable influence for farmer’s cultivation wishes, the
agriculture agency should improve the management efficiency of farms and water resources in addition
to consideration of the hydrological changes and crop yields adjusting.
References
Cheng Chang-Chi, “Study on the Water Resources Distribution by the Adjustment of Rice-Paddy and
Upland-Crop Fields upon the Accession to WTO”, Water Resources Planning Institute , 2006.
Council of Agriculture, “Implementation results of the program of water-resources distribution
concerning the adjustments of rice-paddy and upland crop fields in 2007~2009”, 2011.
Council of Agriculture, “Improve the Efficiency of Irrigation Water Use Planning and Optimization of
Irrigation Management Decision-Making Mechanism”, 2013.
Council of Agriculture, “Reviewing of Reasonable Amount to Agricultural Irrigation Water”, 2012.
Lee Chaur-Shyan, Liu Shin-Yun, “Assessment for the Program of Water-Resources Distribution
Concerning the Adjustments of Rice-Paddy and Upland Crop Fields”, Department of Agricultural
Economics, NCHU , 1999.
Tsai Tsan-Hsiung, “The Impact on Agricultural Water Demand Due to Partial Sabbath of Paddy
Farmin”, Department of Bioenvironmental Systems Engineering, NTU, 1996.
Taoyuan Irrigation Association, ”1995~2001 Statistics Yearbook”, 2002.
Taoyuan irrigation research & development foundation, “Maintenance on the Water Rights and
Investigation on the Water Requirement for Taoyuan Irrigation Association”, 2012.
Water Resources Planning Institute, “Study on the Enhancement of Water Resources Utilization by the
Adjustment of Cultivation on Farm-Lands in Taoyuan Area”, 2005.
THE FUTURE
E-05
Irrigation Practice and Irrigation Management Improvement

in Baingda Irrigation Project
Maung Maung Naing*, Thiha Aung*, Zaw Min Htut*, Yutaka Matsuno**,
Haruhiko Horino***
*Irrigation Technology Centre, Irrigation Department, Bago(Myanmar)
** Department of Environmental Management, Faculty of Agriculture, Kinki University
*** Graduate School of Life and Environmental Sciences, Osaka Prefecture University
ABSTRACT
The Baingda project is a large scale irrigation project in Myanmar and located in Daik-U Township
in Bago Region. The main dam was constructed in 2003 and has a storage capacity of 461 MCM to
irrigate an area of 18900 ha in the dry season, and to protect Pyuntansa floodplain area from flooding
in the rainy season. Traditional rainfed rice is mainly cultivated in the area during the rainy season,
and bean and pulse crops were cultivated as second crop. The irrigation service was started in 2009
for summer paddy in the dry season after the completion of its diversion weir in 2007. This project is
still under construction for irrigation canals. Especially, on-farm irrigation facilities have not been
completely constructed. Thus, water is supplied for irrigation mostly using main and distributary
canals in conjunction with the plot-to-plot system, and about 10% of the area was irrigated last two
seasons. The farmers still continue their traditional system rather than in the irrigation service. So far,
low yield and low income from the new system cause low interest of the farmers. Water supply in main
canal is not so regulated and on-farm irrigation facilities are insufficient. However, this project has
much potential of land and water for increasing of better crop production. In this condition, study and
facilities development are being implemented. This paper discusses the condition of irrigation practice,
the setting of study on secure water supply and balancing, the preparation of facilities development
and the organizing water user groups.
Keywords: Irrigation service, summer paddy, irrigated area, irrigation management,
water balance study, water user groups
1. Introduction
Rice production is a major economy in agriculture sector of Myanmar especially for food
security and increase of export income. Traditional rainfed rice cultivation is common in almost all
areas of the country during the rainy season. However, due to uneven rainfall distribution over the
regions, irrigation is also necessary for supplementary water supply in some regions even in the rainy
season. Furthermore, irrigation is entirely needed for summer paddy cultivation in the dry season.
Thus, to be able to meet the aim, irrigation projects have been implemented throughout the country for
water supply in irrigation. The Baingda Project is one of the large scale irrigation projects in Myanmar
and it was constructed for summer paddy irrigation in the dry season. The project area is located in
Lower Myanmar and it is in one of the major rice production regions and traditional rainfed rice is
widely cultivated. After providing this irrigation system, the farmers have gained better benefit from
rice and other crops production throughout a year. This project is still under the construction of canal
networks and while irrigation service is started, the irrigation facilities are being constructed. The
irrigation management system is still dependant on traditional and conventional practices with low
technical inputs. Especially, the farmers still do not have sufficient experience and knowledge on
irrigation and how it should be applied in their area. Water is supplied for irrigation mostly using main
and distributary (DY) canals in conjunction with traditional plot-to-plot systems and this may cause
difficulties in water control and uneven water distribution. Such a condition has been found in the on-
farm irrigation development study (JIID 2007) and the need for canal facilities and farmers' water user
group development have been discussed (M M Naing 2004a; M M Naing 2005). This project needs to
improve on-farm irrigation facilities and irrigation awareness of farmers. Now-a-days, a few percent of
area is irrigated in the project area. However, this project has much potential of land and water for
increasing of better crop production. In this condition, study and facilities development are being
implemented. This paper discusses the condition of irrigation practice, the setting of study on secure
water supply and balancing, the preparation of facilities development and the organizing water user
groups.
2. The characteristics of Baingda Irrigation Project

The Baingda project is a large scale irrigation project in Myanmar and located in Daik-U
Township in Bago Region in Lower Myanmar. The main dam is earthen dam and was constructed in
2003. The reservoir has a gross storage capacity of 461 MCM with a catchment area of 254 km2 to
irrigate an area of 18900 ha in the dry season, and to protect Pyuntansa floodplain area from flooding
in the rainy season (Design Branch, 2001, Irrigation Technology Centre, 2002). Traditional rainfed
rice is mainly cultivated in the area during the rainy season, and bean and pulse crops were cultivated
after the rainy season as second crop. The irrigation service was started in 2009 for summer paddy in
the dry season after the completion of its diversion weir in 2007. This project is still under
construction for irrigation canals. Especially, on-farm irrigation facilities have not been completely
constructed. Thus, water is supplied for irrigation mostly using main and distributary canals in
conjunction with the plot-to-plot system, and about 10 to 15 % of the area was irrigated last two
seasons.
2.1 Irrigation facilities development stages in Baingda Project

Table 1. Implementation of irrigation facilities development stage in Baingda Irrigation Project
Stage Irrigation Facilities Remark
Stage I Construction of main dam

Stage II Construction of diversion weir
Stage III Construction of main (primary) canal Starting irrigation service
and distributary (secondary) canals (MC
and DY canals)
Stage IV Construction of minor (tertiary) canals On-farm irrigation facilities
Stage V Construction of water courses (WC) and On-farm irrigation facilities and
organizing water user groups (WUGs) improvement of institution
The construction of main dam was given priority and it was constructed in 2003 and then its
diversion weir was constructed in 2007. After that, irrigation canal network is gradually constructed
and developed (Table 1). The stage III was completely constructed in 2012. Before 2012, the
construction of distributary canals was finished up to DY-3 canal. Thus, irrigation service was given to
only in the DY-1 to DY-3 areas. The stage IV and V are still under construction. Especially Minor
canals at the downstream area are still to be constructed. Water courses development and organization
of water user groups mostly remain to be developed in this project area.
2.2 Canal layout of the project

The irrigation network is located only at left side of the Baingda creek. The irrigation water is
supplied mainly by using a main (primary) canal and four distributary (secondary) canals (DY canals)
(Figure 1). After conveying water by these main and DY canals, Minor (tertiary) canals and plot-to-
plot system are applied for on-farm irrigation. Number of Minor the canals organized in each DY
canal is mentioned in Table 2. 36 the Minor canals are used for on-farm irrigation in conjunction with
plot-to-plot system. A few field ditches (water courses) were constructed in this project area by the
farmers.
Baingda dam
Baingda weir
DY-1 Table 2. Water supply system in Baingda

Project
Main DY-2 DY canal Number of Minor Number of Minor

canal canals for water canals for water
supply to left supply to right
DY-1 7 Nos. 5 Nos.
DY-3
DY-2 2 Nos. 11Nos.
Irrigable area DY-3 2 Nos. 5 Nos.
DY-4 3 Nos. 1 No.
DY-4
DY – Distributary , Minor canal – Tertiary canal
Figure 1. Layout of canals in

Baingda Project
2.3 Rainfall and cropping season in the project area

The region is warm and tropical and enjoys the southwest monsoon. The rainfall is abundant
for rice cultivation. Even in drought years, the region receives a stable rainfall for rice cultivation,
with the average rainfall being about 2,500 mm, of which more than ninety percent of the rainfall falls
from May to October. This project area has a rainy season and dry season in a year. The rainy season
is from mid-May to mid-November, and the other half is a dry season. Rivers, rivulets, and natural
drainages are flooding every year during the rainy season due to the monsoon heavy rain. However,
their flow discharges are very limited during the dry season. There is almost no rain during the dry
season (Figure 2). The farmers traditionally cultivate rainfed rice in the rainy season every year and
then beans and pulses crops are cultivated as second crop after the rainfed rice. This practice is still
1200
1000
Monthly rainfall (mm)
800
600 2010
2011
400
2012
200
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
Figure 2. Monthly rainfall in the project area, Daik-U town, (2010, 2011 & 2012)
functioned in the project area although irrigation service has been initiated. Now-a-days, the cropping
pattern has two kinds; one is rainfed rice and upland crops, and another one is rainfed rice and
summer rice (summer paddy) in the project area. Summer rice is totally irrigated by the canal system
in the dry season.
2.4 Irrigated area of the project

So far, the project actual irrigable area decreases to 11500 ha. The irrigated area of summer
paddy was only 1300 ha to 2800 ha within last three seasons from 2010-11 to 2012-13 (Figure 3).
Thus, actual yearly irrigated area is still much smaller than the area available in the project area.
Because, the farmers still continue their conventional ways and original cropping pattern. Summer
rice is irrigated in a few areas. It is seen that the area of upland crops with beans and pulses is more
than the irrigated rice in 2012-13 (Figure 4). The area of upland crops is 1.7 times more than the
irrigated area of summer rice last season. The DY-3 area is mostly irrigated for summer rice compared
with other areas. Among the Minor canal areas of DY-3 canal, the Minor-4 area is the most irrigated
area. Even in this area, upland crops were cultivated about 26% and non cultivated area was about
14 % in the 2012-13 dry season (Figure 4).
3000 6000
2500 5000
Non
Cultivated area (ha)
Irrigated area (ha)
4000 cultivation
2000
3000
1500 Upland
2000 crop
1000
1000
500 Paddy
0
0 DY-3, Minor- Project' total
2010-11 2011-12 2012-13 4 area area
Irrigation season (Year) Location
Figure 3. Irrigated area of the project area Figure 4. Cultivated area in the dry season,
2012-13
Figure 5. Distribution of on-farm plots and the Figure 6. Distribution of plots in elevation in
cultivated area of Minor-4 canal area Minor-4 canal area of DY-3 canal
of DY-3 canal, 2012-13
2.5 The Characteristics of on-farm Irrigation

The distribution of farm plots is still traditional form and the farmers manage water through
plot-to-plot system. They use minor canal (Minor-4 canal) as main delivery canal for their farm plots
(Figure 5). They choose and grow crops according to water availability in the plots. The summer
crops are distributed in the area in the last season (Figure 5). The plots' elevation is also different due
to topographic condition. This different condition causes to use plot-to-plot system in water
distribution and it support field ditch (water course) development (Figure 6). Producing maps and
detail survey for topographic condition is also necessary in this project area. Water courses and field
ditches are necessary for easy water intake and on-farm irrigation management in this project.
2.6 Water release from the reservoir

The irrigation season in the dry season is started from last week of December and ended at
500.00
450.00
Water storage in reservoir (MCM)
400.00
2009-10
350.00 2010-11
300.00 2011-12
2012-13
250.00
Full storage
200.00 Low storage
150.00
100.00
50.00
0.00
10-Mar
20-Mar
30-Mar
10-Jan
20-Jan
30-Jan
10-May
20-May
30-May
10-Dec
20-Dec
30-Dec
10-Nov
20-Nov
30-Nov
10-Apr
20-Apr
20-Apr
10-Feb
20-Feb
30-Feb
Date
Figure 7. Water release from the reservoir for irrigation
about last week of May according to last 3 season from 2010-12 to 2012-13. The reservoir's storage is
gradually drawn down up to 150 MCM for irrigation of summer paddy (Figure 7). The reservoir has
no full storage at the starting time of irrigation. However, some active storage about 80 MCM remains
in the reservoir.
2.6 Water supply in canals

Water supply in canals was measured at main canal, DY-3 canal and Minor-4 canal of DY-3
canal for the 2012-13 season (Figure 8). The operation of water supply in canals is not frequently
done and its water supply is mostly constant except over some period. Especially, water supply in
main canal is almost same throughout the season according to information from the irrigation
maintenance office. In average, it supplies water about 2 m in depth through the season every year.
2.50
Canal water supply in depth (m)
2.00
1.50
1.00
0.50
0.00
14-May-13
21-May-13
1-Jan-13
8-Jan-13
5-Mar-13
15-Jan-13
22-Jan-13
29-Jan-13
7-May-13
12-Feb-13
19-Feb-13
26-Feb-13
12-Mar-13
19-Mar-13
26-Mar-13
16-Apr-13
23-Apr-13
30-Apr-13
5-Feb-13
25-Dec-12
2-Apr-13
9-Apr-13
Irrigation period (Date)

Main canal DY-3 canal Minor-4 cnal
Figure 8. Water supply in canals, 2012-13 season
3. Discussion
1) The Baingda Irrigation Project is still in developing stage. Although main irrigation
facilities were completed, water could not be supplied to all areas of the project. Especially, on-farm
irrigation facilities are necessary for easy and better water supply in irrigation.
2) The farmers do not well know about irrigation. They continue their traditional rainfed rice
cultivation in the rainy season and then they try to grow beans and pulses with low labour and
investment input after the rainy season. They still much rely on their conventional ways and manner
in agricultural works. But, they have to call new challenges in irrigation system.
3) The irrigation office and irrigation engineers should complete irrigation facilities to reach
full stage and complete system. Especially, on-farm irrigation facilities are still in insufficient stage.
Farm ditches (water course) are not mostly constructed. Based on the topographic condition, detail
survey should be done to know elevation of each farm plot. As the result of development work shown
in Figure 6, the detail maps should be improved for all areas and these maps could be used for on-farm
irrigation development. At the same time, Main canal, DY canal and Minor canal should be reviewed
and water supply in these canals should be studied for batter water distribution.
4) The water allocation and balance study is being carried out. It is set up as total project level,
DY canal level and Minor-4 canal level (on-farm level) (Figure 9). This development study will
support the future water allocation and water distribution plan of Baingda Irrigation Project. The
reservoir operation for water allocation will be studied with respect to water resources availability and
on-farm irrigation condition.
Total area (Project whole area)
DY-3 area Minor-4 area
Figure 9. Layout of water balancing study in Baingda Project
5) Producing maps as shown in Figure 6 is very useful for farm ditch development.
Application of GIS technology is being introduced in this project. Such kind of technical input will
bring this project under proper irrigation and improvement stage.
6) The farmers have insufficient knowledge in irrigation. Irrigation awareness training should
be implemented and water user groups should be formed. In this stage, village level administrative
Chairman
Secretary
Members Members Members

Village (i) Village (ii) Village
F & NF F & NF F - Farmers

F & NF
NF - Non-farmers
Figure 10 The organization of a Village Tract in administrative body
system is very important in Myanmar (Figure 10). The village administrative members play in
important role in organizing farmers and non farmers in villages (M M Naing, 2010). Thus, irrigation
training should be also implemented for these members. Now, such trainings are being given to the
farmers and village administrative members to improve irrigation in the project area.
4. Conclusions
1) The traditional system of rice production still has much impact on possibility of irrigation
service improvement. The cropping pattern of the region should be improved by using the irrigation
facilities of Baingda Irrigation Project (M M Naing2004b). More effective practice and operation of
Baingda Irrigation Project could improve the socio-economic condition of the farmers there.
2) On-farm irrigation development is necessary in this project area. Providing better irrigation
facilities could improve the farmers' interest in irrigation and project's optimum benefit.
3) Education and training of irrigation awareness should be lectured both for farmers and
departmental persons concerned as well. Farmers' water user groups should be established and
organized for better irrigation development and management in this area.
4) Monitoring and development study should be carried out for better crop production and water
management improvement in the project.
Acknowledgement
The authors express their gratitude to the Director General of the Myanmar Irrigation
Department for his support to this collaborative study of the Irrigation Technology Centre (ITC) and
Japanese Consortium of Irrigation Research (JCIR). To present this content on the PAWEES 2013
conference, the financial and necessary support was provided by the collaborative study program.
References
Design Branch (2001) Specification of Earth & Concrete Works for Baingda Dam Project, Irrigation
Department, Yangon
Irrigation Technology Centre (2002), Report on Quality Control Work of Baingda Irrigation Project
(in Myanmar), Irrigation Department, Bago
Maung Maung Naing (2004a) Towards participation in adaption of the technical measures for water
resources management, Myanmar Engineering Society, CAFEO-22 paper 628, 01-09
Maung Maung Naing and Satoh, M. (2004b) Effective use of a reservoir for paddy irrigation in
tropical monsoon Asia – a case study of the Ngamoeyeik Project, Lower Myanmar –, Paddy and
Water Environment, Vol.2, No.1, 19-25.
Maung Maung Naing (2005) Paddy Field Irrigation Systems in Myanmar (Country paper) The Future
of Large Rice-Based Irrigation Systems in Southeast Asia, (Regional Workshop, 26–28 October,
2005, Ho Chi Minh City, Viet Nam,), FAO Regional Office for Asia and the Pacific, Thailand
JIID (2007) Guidelines for On-farm Irrigation Development and Management in Monsoon Asian
Countries, Japanese Institute of Irrigation and Drainage (JIID), Tokyo, Japan
Maung Maung Naing (2010) Conditional Assessment of Farmers’ Participation in Irrigation
Management in Myanmar, Proceedings of the First Symposium on Water Resources
Development and Management in Irrigation Projects, Irrigation Department
THE FUTURE
E-06
Nitrogen and Weed Management in No-tilled Transplanted Rice

on No-tilled Transplanted Rice- Surface Seeded Wheat Cropping
System under Conservation Agriculture
Pijush K Mukherjee , Biswapati Sinha

Department of Agronomy, Uttar Banga Krishi Viswavidyalaya (UBKV), Pundibari Cooch
Behar736165, West Bengal, India (Email:pkm_agronomy@yahoo.co.in)
ABSTRACT
Rice is generally cultivated under puddled condition which in long run destroys soil properties and
thus ultimately affects sustainability of production system. Conservation Agriculture (CA) has the
potentiality to restore sustainability of production system. High rainfall during onset of monsoon
creates narrow window for dry direct seeding of rainy season rice, however, sandy loam soil and
stagnant water creates an opportunity for no tilled transplanted rice. Weed management
encompassing brown manuring, maize residue mulching, hand weeding and weedy control were laid
out in main plot. N management like recommended N (150 kg N ha-1 with 30 kg N ha-1 as basal and
three split @ 40 kg N ha-1 at primary tiller initiation, active tillering and panicle initiation stages),
vermicompost based on 30 kg N ha-1+ N top dressing through LCC (LCC 4 value @ 30 kg N ha-1) and
basal N (30 kg ha-1)+N top dressing through LCC were assigned in subplot of split-plot design with
three replication in hybrid rice (Arize 6444). Practice of green manuring with Sesbania rostrata after
wheat followed by its use as surface mulch after harvest, conserving rain water by maintaining high
bund around the field and crop residue management facilitated easy execution of manual rice
seedling transplanting under no tilled condition. Rice plant receiving N management through LCC
generated demand for N at 28 (Active tillering stage), 42 (Just before the initiation of panicle
primordial) and 63 (Beginning of booting stage) days after transplanting and reduced N requirement
from 150 kg ha-1 to 120 kg ha-1 a saving of 30 kg N ha-1. Vermicompost application delayed the final
N demand up to final booting stage or beginning of heading stage. Brown manuring showed highest
value of weed control efficiency because of weed suppression capacity of Sesbania rostrata coupled
with 2,4-D application. Crop residue management and maintaining stagnant rain water facilitated
easy execution of manual transplanting in no tilled condition. Site specific and real time N nutrition
through LCC improved rice productivity and saved certain amount of N.
1. Introduction
The rice-wheat cropping pattern of South Asia has seen a phenomenal growth in area,
production and yield in the last two decades. This cropping system encompasses the four
countries of Bangladesh, India, Nepal and Pakistan along the Indo-Gangetic Plains (IGP) and
into the mid-hills of the Himalayas. With the introduction of improved high yielding, input
responsive, short duration rice and wheat varieties, the rice-wheat pattern became feasible
and it was possible to grow both the crops grown in the same calendar year. Rice (Oryza
sativa L.)–wheat (Triticum aestivum L. emend. Fiori & Paol) is the most popular cropping
system of India. It is practised on 24 million ha. In India, the area under this rotation is about
11.0 million ha; mainly concentrated in Indo-Gangatic Plains. Rice–wheat cropping system is
very productive but it has fatigued natural resources resulting in decline in groundwater level,
soil carbon stocks, and soil plant available nutrients and buildup of pest and diseases (Gupta
et al. 2006). Delay in sowing of rice due to unavailability of machine for puddling operation,
delay in onset of monsoon, unavailability of manual labour for transplanting operation
resulted in difficulties of raising succeeding wheat crop because of short turnaround time,
huge residue load of rice generated through combined harvester operation, high weed
pressure during critical period of weed-crop completion and finally exposure of terminal
growth stage of wheat to the warm weather condition. Burning residue for clearing land and
intensive tillage practices caused loss of nutrients especially valuable micro-nutrients and
over exploitation of petrol and diesel, the valuable natural resources. These events appeared
in addressing the issue of conserving natural resources in mechanized farming and side by
side improving system productivity. The most common practice for establishing rice in the
rice-wheat (Triticum aestivum L. emend. Fiori & Paol.) system of Indo-gangetic plain region
is puddling before transplanting. The alternative to puddling could be direct seeding because
it does not require that heavy amount of labour, water and capital input initially and also crop
mature earlier (7-10 days) than the transplanted crop allowing timely sowing of succeeding
wheat crop (Giri 1998). Recent research suggests that new method of rice establishment, viz
zero till rice, bed planting has potential to reduce costs and increase sustainability of irrigated
rice culture while maintaining yield (Hobbs and Gupta 2002). Conventional method of wheat
sowing requires intensive pre-planting cultivation, which are labour, time and energy
intensive. To make this practice, easier and cost effective, several crop establishment
methods, viz zero till drill, roto drilling, strip till drill and surface seeding of wheat etc. have
been developed (Yadav et al. 2006). Zero tillage with rice stubbles retention may improve the
soil physico-chemical condition by improving the organic matter in the soil. Therefore, these
issues led to the need of exploring alternate crop establishment techniques for restoring
sustainability of rice-wheat cropping system, maintaining and conserving natural resources.
Conservation Agriculture has the potentiality of restoring sustainability of the production
system by preventing overexploitation and degradation of natural recourses, reversing the
process of exploitation and degradation to regeneration and conservation of natural recourses
and maintaining quality of environment, which ultimately leads to optimization of biosphere
for agricultural production. In the last 120 years, intensive agriculture has caused a carbon
loss between 30 and 50% and over the past 150 years the amount of CO2 in the atmosphere
has increased by 30%. Since the mechanization of agriculture began a few hundred years ago,
scientists estimate that some 78 billion tonnes of carbon once trapped in the soil have been
lost to the atmosphere as CO2 (Lal et al. 2004). In the present day of intensive agriculture CA
has been seen as a stabilizer of the production system while mobilizing resource degradation
to resource generation and conservation. Success of CA is largely influenced by crop
establishment technique. Therefore, caution should be taken to avoid blanket adoption of crop
establishment techniques used for CA. It should be site specific and it should have need based
approach (Bhale and Wanjari 2009). Terai region of West Bengal, India has the
characteristic feature of high rainfall during the rainy season. Long term data analysis of
rainfall shows that rainfall in this region starts during the first fortnight of April and later
merges with the onset of monsoon usually appears during first fortnight of June. Continuous
rainfall starting from April followed by heavy down pour during first fortnight of June creates
very narrow window for dry direct seeding of rice. Previous experiments carried out in these
areas on zero tilled direct seeded rice showed that continuous rain and extreme rainfall events
made dry direct seeding highly vulnerable in terms of its implementation and successful
establishment of rice after seeding. Rainfall pattern of these areas is a major bottle neck of
blanket adoption of dry direct seeding of rice and that lead to the need of capitalizing the
usefulness of climatic condition and finding appropriate alternate crop establishment
technique to be supporting one of the guiding principles of conservation agriculture i.e.
minimum or no tilling of soil. Nature of soil texture and stagnant water due to huge down
pour during the onset of monsoon creates an opportunity of exploring possibility of no tilled
transplanted rice as an alternative crop establishment technique against puddled transplanted
rice. Because of the high rainfall during monsoon period, low land situation contains high
residual soil moisture after harvesting of rice. This high residual soil moisture makes the land
inaccessible to the implements for the purpose of land preparation and this in turn resulting in
difficulties of raising winter crops. Soil moisture coming down to the level of workable
condition causes considerable delay in sowing beyond the optimum time of most of the
winter crops. Majority of the farmers usually keep these lands uncultivated or poorly
managed with surface seeded lathyrus. However, potentiality of high residual soil moisture
needs to be harnessed for successful establishment of wheat, the important winter crop in
dominated rice-wheat cropping system. This region also has high ecological land use pattern
of wheat because of prolonged winter and prevalence of low temperature during grain filling
stage which is an ecological nitch of wheat cultivation. It is worthwhile to mention here the
prevalence of prolonged winter, the high residual soil moisture is conduct to augment the
yield of wheat. Lower yield rate recorded for wheat despite its natural advantage, can
therefore, largely be attributed to the lower level of use of key inputs like fertilizer, irrigation
etc. A wide spread decline in yield and factor productivity have been noted in rice (Oryza
sativa L.)-wheat (Triticum aestivurn L. emend. Fiori & Paol) cropping system (Ladha et al.
2003). As a consequence, farmers resorted to use greater than recommended rates of fertilizer
N to maintain the previously attained yield levels with relatively less fertilizer (Dwivedi et al.
2001) which resulted in lower nitrogen (N)- use efficiency. Increase in fertilizer N application
rate may also enhance the extent of NO,-N leaching and thereby pollution of groundwater
(Singh et al. 1995). As the blanket N recommendations do not consider the variability in
indigenous N supply and location specific need for crop N demand and thus efficient use of
fertilizer N is further restricted (Adhikari et al. 1999). The leaf color chart (LCC) is an
innovative cost effective tool for real-time or crop-need-based N management in rice, maize
and wheat. The need based N management in hybrid rice using LCC has the potential of
replacing the blanket uniform fertilizer rates recommended across vast areas. It has already
been found very useful in efficiently managing fertilizer N in inbred rice cultivars (Singh et
al. 2007a, Singh et al. 2007). Thus, it is an eco-friendly tool in the hands of farmers. Keeping
the fact in view the experiments have been conducted with the objectives to study the weed
and nitrogen management in no-tilled transplanted rice and its effect on no tilled surface
seeded wheat.
The field experiments were conducted during 2010-11 and 2011-12 at research farm
of Uttar Banga Krishi Viswavidyalaya, Pundibari, Cooch Behar, West Bengal, India. The
place is located in terai region of West Bengal at 25057' N to 270 N latitude and 88025' E to
89054' E longitude at an elevation of 43 meters above mean sea level and the land topography
is low land. The soil of the experimental site was sandy loam with sand 62.84%, silt 20.48%
and clay 15.92%. Nitrogen, phosphorus and potassium content of the soil were 112.35, 15.63
and 89.22 kg/ha, respectively. The treatments consist weed management as main plot factor
encompassing brown manuring, maize residue mulching, hand weeding and weedy control. N
management like recommended N (150 kg N ha-1 with 30 kg N ha-1 as basal and three split @
40 kg N ha-1 at primary tiller initiation, active tillering and panicle initiation stages),
vermicompost based on 30 kg N ha-1+ N top dressing through LCC (LCC 4 value @ 30 kg N
ha-1) and basal N (30 kg ha-1)+N top dressing through LCC were assigned in subplot of split-
plot design with three replication in hybrid rice (Arize 6444). Before the experimentation of
rice, Sesbania rostrata (seed rate of 40 kg ha-1) was grown in no tilled surface seeded
condition, however, during the 2nd year of experimentation the Sesbania was surface seeded
within the anchor residue of wheat. 40 days old Sesbania plants were cut at the base of the
plant and were retained on surface soil. The land preparation for rice was done only by
controlling existing weed flora with the use of glyphosate at the dose of 2.0 kg ha-1 as pre-
plant desiccators. The rice (Aize 6444-hybrid rice) seedlings were raised in conventional
nursery. The main field was surrounded by bund which was finally made into raised bund 50
to 60 cm height. This high bund was made with the purpose of conserving rain water as well
as maintaining submergence. After heavy rain the stagnant water for the period of 3 to 5 days
made the soil soft and ideal for transplanting. Decomposition of surface residue of Sesbania
rostrata plants and anchor residue of preceding wheat in stagnant water facilitated easy
execution of transplanting of rice seedling like puddled condition. The concept of brown
manuring as weed management practice was tested, in which Sesbania rostrata (seed rate of
40 kg ha-1) was grown at inter row and intra row spaces. In Sesbania rice co-culture, Sesbania
was killed with the application of 2,4-D 0.5 kg ha-1 at 30 days after sowing (DAS) of
Sesbania and then created mulching with the help of paddy weeder. Site specific and real
time N management was made through leaf colour chart (LCC). The topmost, fully expanded
healthy leaves of each of the 10 plants were compared with the LCC at 7 days interval
usually starting from 15 days after transplanting (DAT) (Gupta et al. 2011). Middle part of
the leaf without detaching from the plant was placed on top of the LCC’s colour strips for
comparison. The readings were taken at same time of the day (8-10 AM) without exposing
the LCC to direct sunlight. Phosphate (60 kg P2O5/ha) was applied as basal and potash (60 kg
K2O/ha) was applied in rice in two split doses. 2/3rd of total amount was applied as basal and
remaining 1/3rd was applied during final top dressing of nitrogen i.e. panicle initiation stage
(45 DAT). Nitrogen, phosphorus and potassium were applied in the form of urea, single super
phosphate (SSP) and muriate of potash (MOP).
3. Result
3-1. Weed flora

The weed flora recorded in no-tilled transplanted experiment were different from the
weed flora generally appears in puddled transplanted rice. Weed flora like Cynodone
dactylon, Cyperus rotundus, Cyperus flavidus, Ageratum conyzoides (during the early
developmental stages) and Spilanthes paniculata were appeared in no-till system, however,
these weeds generally do not appear in puddled system. Sedges comprised by Cyperus
rotundus, Cyperus flavidus, Cyperus iria, Fimbristylis miliacea constituted the dominant
weed flora at the different developmental stages of rice. Among the sedges, emergence of
Cyperus rotundus was rapid during early developmental stages which was followed by the
emergence of Cyperus iria, Fimbristylis miliacea and Cyperus flavidus. The broadleaved
weed Ludwigia parviflora was the dominant weed since the active tillering stage to the
maturity stage of rice. During the initial developmental stages up to 40 days after
transplanting (DAT), continuous emergence of Cyperus rotundus supported increased
population of sedges, however, it had narrow emergence profile, whereas the other sedges
like Cyperus iria, Fimbristylis miliacea, Cyperus flavidus had longer emergence profile. The
broadleaved weed Ageratum conyzoides had higher rate of emergence during the initial
developmental stages up to 40 DAT resulted in steep increased in broadleaved weed
population, however, mortality of Ageratum conyzoides beyond 40 DAT caused stagnation of
broadleaved weed population which was later increased due to higher rate of emergence of
Ludwigia parviflora and emergence of other broadleaved weeds like Spilanthes paniculata
and Enhydra fluctuans in which Ageratum conyzoides had narrow emergence profile whereas
Ludwigia parviflora had long emergence profile. Among the grasses Cynodon dactylon and
Echinochloa colonum had long emergence profile resulted in increasing trend of population
at the different development stages and contributed higher proportion of total weed
population in comparison to the broadleaved weeds. Similar finding was reported by Singh et
al. 2011.
3-2. Weed control efficiency

Introduction of Sesbania rice co-culture controlled the weeds effectively as it acted as
a smoother crop to suppress the growth of the weeds. Brown Manuring registered highest
values of weed control efficiency at different developmental stages ranging from 56.70 to
78.14 % in 2010-11 and 63.53 to 80.10 % in 2011-12. Similar finding was reported by
Ravisankar et al. 2008 and Joseph et al. 2008. High growth rate of Sesbania resulted in
successful competition with several grasses, sedges and broadleaf weeds since early growth
stages of rice. Killing of Sesbania with 2,4-D at 30 days after sowing followed by its
incorporation through paddy weeder made duel effect in the system in terms of controlling
sedges and broadleaves by 2,4-D as well as mechanical control of the weeds those have
escaped the action of 2,4-D and turning the Sesbania as manure as well as creating mulch
which prevented emergence of weeds especially broadleaves. Maize residue as surface mulch
also shown the capacity in suppressing growth of the weeds through the way of preventing
weed seeds germination. All the nitrogen management treatments did not show much
variation on weed control efficiency.
3-3. Nitrogen management:

In case of blanket recommended N application N is applied in 4 splits i.e. ¼ as basal,
¼ as first top dressing at 15 days after transplanting (primary tiller initiation stage), ¼ as
second first top dressing at 25 days after transplanting (active tillering stage) and ¼ as 3rd top
dressing at 45 days after transplanting (Initiation of panicle primordia). In each split, N was
applied at the rate of 37.5 kg ha-1. However, in case of N management through LCC it was
observed that rice plant receiving basal treatment of 30 kg N ha-1 generated demand for N at
28, 42 and 63 days after transplanting which were corresponding to active tillering stage, just
before the initiation of panicle primordia stage and booting stage, respectively. Similar
finding was reported by Gupta et al. 2011. In each stage N was top dressed at the rate of 30
kg ha-1 and total N requirement was reduced from 150 kg ha-1 to 120 kg ha-1 a saving of 30 kg
N ha-1. In case of vermicompost+N management through LCC treatment, rice plant created N
demand at 21, 42 and 70 days after transplanting which were corresponding to beginning of
active tillering stage, just before the initiation of panicle primordial stage and final booting
stage or before the heading stage, respectively. In each stage N was top dressed at the rate of
30 kg ha-1 and total N requirement was reduced from 150 kg ha-1 to 90 kg ha-1 as 30 kg N ha-1
was applied in the form of vermicompost as basal. Result of LCC based N management
reveals that LCC tool enabled to judge the timing of N application and variation of timing
was mainly due to the availability of N from soil. In vermicompost+N management through
LCC rice plant created demand after transplanting at the beginning of active tillering stage
instead of primary tiller initiation stage which is generally followed as blanket
recommendation. In case of basal 30 kg N ha-1+N management through LCC treatment,
demand for N was delayed up to active tillering stage as N was available due to its basal
application. Vermicompost first release the N slowly and plant created N demand earlier than
basal 30 kg N ha-1+N management through LCC treatment. However, at later growth stages
of rice, vermicompost delayed the final N demand up to final booting stage or beginning of
heading stage and N nutrition at this stage is very important as it increases the longevity of
flag leaf which is a main contributor of photosynthates during the grain filling stage. In
blanket recommendation the final top dressing of N is only made during the initiation of
panicle primordia (PI) and N application beyond this stage encourage elongation at lower
most internodes, which in turn, increases bending moment and reduce breaking strength of
the rice plant and ultimately makes the rice plant susceptible to lodging. However, it was
observed that top dressing of potassium (1/3 of total amount) along with final top dressing of
N beyond PI stage maintained the stability of rice plant in both the treatments basal+N
through LCC and vermicompost+ N management through LCC. Development of female
panicle (6 to 8%) was observed in vermicompost+LCC treatment. Development of female
panicle indicates good N nutrition in rice and this was made through vermicompost+ N
management through LCC treatment. Vermicompost slowly released N in the soil and LCC
measured the real time demand of rice for N which was met by top dressing of nitrogenous
fertilizer (urea). This treatment combination facilitated site specific and real time N
management for rice. In addition to this advantage some kind of growth promoting activities
was observed on rice in vermicompost treated plots resulting in higher plant height.
3-4. Yield attributes and grain yield

Among the different levels of main plot factor the result on grain yield revealed that
there was no significant difference on grain yield between brown manuring and hand
weeding. This was mainly due to better weed control, which in turn, minimise the magnitude
of weed-crop completion leading to better values of yield attributes. The mulching treatment
registered higher tiller number, however, more proportion of unfilled grain development
reduced ultimate yield of the mulching treatment which was significantly lower than hand
weeding treatment in 1st year, however, in second year the yield was at par with hand
weeding treatment. The magnitude of yield attributing characters during the 1st year of
experimentation was comparatively higher than the yield attributing characters obtained
during 2nd year of experimentation, however, final yield obtained during 2nd year was
comparatively higher than that of 1st year experimentation. This was because of down pour
for a short period of time during the harvesting time of 1st year experimentation (Table 1).
3-5. Nitrogen management

In subplot factor the vermicompost+ N management through LCC treatment
registered highest yield and this was due to effective N nutrition through LCC technique and
slow release of N from the vermicompost. The development of female panicle also helped in
achieving highest yield.
3-6. Interaction between weed management and nitrogen management

Among the interaction effects highest value of grain yield registered in the treatment
combination Hand weeding + Vermicompost with N management LCC (5.54 tones ha-1 in
2010-11 and 5.87 tones ha-1 in 2011-12) which was closely followed by treatment Brown
Manuring + Vermicompost with N management through LCC (5.50 tones ha-1 in 2010-11
and 5.76 tones ha-1 in 2011-12) without having any significant difference among each other
which was mainly due to synergistic effect of effective weed control and real time N nutrition
to the crop. Effectiveness of the treatment combination brown manuring+vermicompost+ N
management through LCC was comparatively higher if assessment could be made in term of
engagement of manual labour for hand weeding.
3-7. Economics:
Highest gross income was obtained in hand weeding+vermicompost+N management
through LCC treatment (Rs 60678/$979 and Rs 68989/$1112 ha-1 in 2010-11 and 2011-12,
respectively), however, net return of this treatment was comparatively lower than other
treatment combinations because of high input cost incurred due to engagement of manual
labour for hand weeding and cost of required vermicompost (Rs 7,700/$124 ha-1). Low net
return ultimate reduced the profit of the treatment in terms of net return per rupee invested
(0.48 and 0.75 in 2010-11 and 2011-12, respectively). Highest net return (Rs 31891/$514 and
Rs 40031/$646 ha-1 in 2010-11 and 2011-12, respectively) and net return per rupee invested
(1.31 and 1.76 in 2010-11 and 2011-12, respectively) was obtained in mulching with maize
residue+ N management through LCC and this was due to low input cost incurred through
weed control and N management by LCC. All the weedy control treatment combination
recorded lowest gross income, net return and net return rupee invested because high weed-
crop competition resulting in lower grain yield. Negative value of net return (Rs -5973/$-96
ha-1 in 2010-11) was obtained in weedy control +Vermicompost+ N management through
LCC treatment and this was due to high cost of vermicompost (Rs 7,700/$124 ha-1) and low
grain yield obtained in the treatment. Higher net return obtained during 2nd year was due to
higher minimum support price (Rs 800/$13 more tone-1) and higher grain yield.
4. Conclusion
Worldwide conservation agriculture (CA) plays a major role in sustainable

agricultural production. It has the potential to emerge as an effective strategy to the
increasing concern of serious and widespread natural resources’ degradation and environment
pollution, which accompanied the adoption and promotion of green revolution technologies.
CA holds tremendous potential for all sizes of farms and agro-ecological system, but its
adoption is probably most urgently required by smallholder farmers. Crop residue
management (CRM) in no-tilled rice-wheat system hold the promise of the adoption of CA
based agricultural practices by small land holder. Maintaining stagnant rain water and surface
residue created easy execution of transplanting in no tilled condition. Site specific and real
time N nutrition through LCC improved rice productivity and saved certain amount of N.
Mulching with maize residue, anchor residue of respective crops and green manuring with
Sesbania rostrata has turned out as an effective crop residue management for controlling
weed, addition of residue and contribution of NPK in soil. No tilled rice-wheat cropping
system provided acceptable profit and therefore, it has the potential to be adopted by small
land holder as CA based agricultural practices.
Table 1. Effect of treatment combinations on weed control efficiency (%) at 60 DAT, yield
attributes, grain yield and net return.
Treatment Weed No of tillers No of filled Test weight Yield Net return Net
control (m-2) grains (g) (tone ha-1) (Rupees/Rs return
efficiency panicle-1 ha-1) (Dollar/$
(%) at 60 ha-1)
DAT
Y1 Y2 Y1 Y2 Y1 Y2 Y1 Y2 Y1 Y2 Y1 Y2 Y1 Y2
Brown 77.38 80.73 406.66 405.00 154.34 144.55 22.21 21.43 5.18 5.41 30889 39442 498 636
manuring +
Recommended N
Brown 72.02 74.18 430.00 420.33 152.31 141.59 21.46 21.22 5.50 5.76 27191 36331 439 586
manuring +
Vermicompost
with N through
LCC
Brown 64.73 66.72 441.66 430.66 151.47 141.51 23.55 21.27 5.22 5.56 31665 41536 511 670
manuring + N
through LCC
Mulching with 61.63 63.73 396.33 388.33 136.20 128.10 20.14 19.66 5.03 5.25 30683 38987 495 629
maize residue +
Recommended
N
Mulching with 61.61 63.79 424.66 416.33 134.30 127.94 21.94 18.55 5.20 5.52 24044 33627 388 542
maize residue +
Vermicompost
with N through
LCC
Mulching with 61.50 63.96 420.33 411.00 135.59 127.72 22.45 18.72 5.13 5.33 31891 40031 514 646
maize residue +
N through LCC
Hand weeding + 100 100 389.33 380.66 158.73 150.11 22.38 21.67 5.22 5.38 24416 32175 394 519
Recommended
N
Hand weeding + 100 100 401.00 390.33 159.00 149.58 23.28 22.31 5.54 5.87 19557 29548 315 477
Vermicompost
with N through
LCC
Hand weeding + 100 100 394.66 385.00 161.22 148.62 22.75 21.36 5.32 5.65 25750 35586 415 574
N through LCC
Weedy control + 0.00 0.00 293.66 281.66 94.77 86.39 20.55 19.30 1.99 2.53 -37 9942 -1 160
Recommended
N
Weedy control + 0.00 0.00 290.33 279.33 94.24 86.12 19.40 19.41 2.21 2.85 -5973 5220 -96 84
Vermicompost
with N through
LCC
Weedy control + 0.00 0.00 285.33 277.00 92.20 86.09 18.67 19.14 2.07 2.64 180 10463 3 169
N through LCC
LSD (0.05) 0.8 0.7 14.1 13.8 2.6 1.8 0.5 0.2 0.13 0.11 - - - -
DAT - Days after transplanting; LCC-Leaf Colour Chart, Y1= 2010-11, Y2= 2011-12
References
Adhikari, C, Bronson, K. F., Panuallah, G. M., Regmi, A. P., Saha, P. K., Dobermann, A., OIk, D. C.,
Hobbs, P. R., and Pasuquin, E., 1999, “On-Farm N supply and N nutrition in the rice-wheat
system of Nepal and Bangladesh”. Field Crops Research 64: 273-86
Bhale, V.M., and Wanjari, S.S., 2009, “Conservation Agriculture: A new paradigms to increase
resource use efficiency”. Indian Journal of Agronomy 54 (2): 167-177.
Dwivedi, B. S., Shukla Arvind K., Singh, V. K., and Yadav, R. L., 2001, “Results of participatory
diagnosis of constraints and opportunities (PDCO) based trials from the state of Uttar Pradesh”.
(in) Development oJ Farmers' Resource Based Integrated Plant Nutrient Slipply System:
Experience of FAO- ICAR-IFFCO Collaborative Project and AICRP on Soil Test Crop Response
Correlation. Subba Rao and Srivastava S (Eds).IISS, Bhopal, India. pp 50-75.
Giri, G. S., 1998, “Effect of rice and wheat establishment techniques on wheat grain yield”. (in)
Proceedings of Rice- Wheat Research Project Workshop, pp 65-68. Hobbs P R and Ray Bhandari
(Eds).
Gupta, R., Jat, M. L., Singh, S., Singh, V. P., and Sharma, R. K., 2006, “Resource conservation
technologies for rice production”. Indian Farming 56 (7): 42–5.
Gupta, R. K., Singh, V., Singh, Y., Singh, B., Kumar, A., and Vashistha, M., 2011, “Need-based
fertilizer nitrogen management using leaf colour chart in hybrid rice (Oryza sativa)”. Indian
Journal of Agricultural Sciences 81 (12): 1153–7.
Hobbs, P. R., and Gupta, R. K., 2002, “Rice-wheat cropping systems in Indogangetic plains. Issues of
water productivity in relation to new resource conserving techniques”. (in) Proceedings of the
Water Productivity Workshop, held during 12- 14 November 2001 at IWMI, Sri Lanka.
Joseph, M., Rajendran, P., and Hemalatha, M., 2008, “Nitrogen levels and green manure
intercropping on growth analysis of wet seeded rice”. Environment and Ecology. 26 (1A): 356-
360.
Ladha, J. K., Pathak, H., Padre, A. T., Dawe, D., and Gupta, R. K., 2003, “Productivity trends in
intensive rice-wheat cropping systems in Asia”. (in) Improving the Productivity and
Sustainability of Rice-wheat Systems: Issues and Impact, pp. 45-76. Ladha J K et al. (Eds).
Special Publication 65, Agronomy Society of America ASA, CSSA. and SSSA. Madison.
Lal, R. M., Griffin, J., Apt, L. Lave, and Morgan M. G., 2004, “Managing soil carbon”. Science
304:393.
Rawat , S. N., and Verma, M. R., 2006, “Performance evaluation of zero till fertilizer seed drill for
wheat crop”. Karnataka Journal of Agriculture Science 19(2): 348–51.
Ravisankar, N., Chandrasekaran, B., Raja, R., Din, M., and Chaudhuri, S. G., 2008, ”Influence of
integrated weed management practices on productivity and profitability of wet seeded rice (Oryza
sativa)”. Indian Journal of Agronomy. 53 (1): 57-61.
Singh Varinderpal, Singh Yadvinder, Singh Bijay, Singh Baldev, Gupta, R. K., Singh Jagmohan,
Ladha, J. K., and Balasubramanian, V., 2007, “Performance of site specific nitrogen management
for irrigated transplanted rice in northwestern India”. Archives of Agronomy and Soil Science 53:
567–79.
Singh Yadvinder, Singh Bijay, Ladha, J. K., Bains, J. S., Gupta, R. K., Jagmohan, Singh, and
Balasubramanian, V., 2007a, “On-farm evaluation of leaf color chart for need-based nitrogen
management in irrigated transplanted rice in north-western India”. Nutrient Cycling in
Agroecosystems 78:167–76.
Singh, Y., Singh, V. P., Singh, G., Yadav, D. S., Sinha, R. K. P., Johnson, D. E., Mortimer, A. M.,
2011, “The implications of land preparation, crop establishment method and weed management
on rice yield variation in the rice-wheat system in the Indo-Gangetic plains”. Field Crops
Research 121 (1) 6474.
Yadav, D., Sushant, S., and Yadav, V., 2006, Alternate crop establishment methods in rice (Oryza
sativa)-wheat (Triticum aestivum) cropping system. National Symposium on Conservation
Agriculture and Environment, held during 26–28 October 2006, BHU, Varanasi.
Session 6
THE FUTURE
F-02
Modeling the Future Water Footprint of Paddy Rice in the

Republic of Korea
Temba Nkomozepi , Sang-Ok Chung

Department of Agricultural Civil Engineering, Kyungpook National University,
Daegu 702-701, Republic of Korea.
ABSTRACT
The vulnerability of agriculture to the rising temperatures coupled with altered rainfall patterns that
occur as a result of climate change is of major interest in agricultural research. In this study, the
impact of climate change on the water footprint (WF) (m3 t-1) of paddy rice in the 2021 – 2040 (2030s),
2051 – 2070 (2060s) and 2081 – 2100 (2090s) was modeled and assessed for the Republic of Korea.
The future crop water requirement, effective rainfall and irrigation requirement from which the WF is
derived were computed using a water balance method and climate projections from 12 General
Circulation Models (GCMs) for three Representative Concentration Pathways (RCPs). The Water
Accounting Rice Model (WARM) was calibrated and then applied to simulate paddy rice yield. The
paddy yield was predicted to decrease by up to 40% by the end of the century. The WF was predicted
to significantly increase particularly in the Gyeongbuk Province. The spatial and temporal uncertainty
of the WF were investigated and found to vary and be the highest for the 2090s and RCP8.5 in the
Gyeongbuk and Jeonam Provinces. The ratio of the blue to the green WF was also predicted to be
altered in the future. We found that some provinces could be less suitable for paddy rice cultivation in
the future and future research should be carried out to facilitate for climate change adaption.
Key words: Paddy rice, climate change, CROPWAT, water footprint, RCP, WARM
1. Introduction
The competition in the supply of sufficient water of good quality for agriculture, aquatic life,
wildlife refuges, recreation, scenic values, riparian habitats, municipal and industrial uses is
exacerbated by increasing water pollution and climate change (Bouwer, 2000). For agricultural
managers, a multi-disciplinary approach to integrated water resource management which incorporates
the balancing of green and blue water flows in agriculture is required to cope with changes in
availability and demands for water (Rockström et al., 2004). In addition to affecting crop water supply
and consumption climate change will also affect crop yields. Agricultural production is vulnerable to
the changes in greenhouse gas concentrations that cause shifts in the rainfall, radiation and
temperature patterns (Supit et al., 2012). Quantifying and understanding the dynamics of the biomass
produced per amount of water used at the crop level is essential for facing agro-economic and
environmental challenges (Tallec et al., 2013). Integrated assessments of interdisciplinary content that
couple scientific and policy issues have been used to investigate the potential climate change and its
impacts in strategic planning (Van Minnen et al., 2000).
Recently, the volumetric and the life cycle analysis approaches to the water footprint (WF) has
been used in the assessment of water resources utilization during agricultural production to improve
agricultural water management practices (Sun et al., 2013a; Yoo et al., 2013a; Bocchiola et al., 2013).
Conceptually, the WF (i.e. the volume of fresh water that is consumed per weight of produced goods)
of paddy rice can be partitioned into blue (surface or groundwater consumed) and green (effective
rainfall consumed) consumptive water use components (Hoekstra, 2003).
Crop yield and water demand response to climate change can be assessed using simple
regression approaches. However, the application of more complex process-based crop simulation
models is justified by their additional capacity to analyze the dynamic interactions between
environment, genotype and management factors (Angulo et al., 2013). Crop models are able to
simulate measured grain yield and other factors with accuracy under diverse environments if input
information is sufficient (Asseng et al., 2013).
Rice yields have significantly increased over the last century as a result of better management of
pests and diseases, genetic improvement and higher applications of fertilizer (Semenov et al, 2012). In
the Republic of Korea (Korea hereafter), the average ambient temperature and rainfall increased by
1.5°C and 10 %, respectively, between 1904 and 2000 (Yoo et al., 2013b). In the event of accelerated
increase in temperature caused by climate change, the rates of grain filling and leaf senescence will
increase while the durations of grain filling and leaf senescence will decrease, thereby decreasing
yields (Kim et al., 2011). Simultaneously, the increase in the ambient temperature will exacerbate crop
water demand and moisture losses associated with water storage and conveyance (Mehta et al., 2013).
Chung (2013) estimated that the total volumetric irrigation water demand will increase by at least 5%
by the end of this century.
Producing more biomass per unit area with less water is of particular interest for rice (Oryza
Sativa) production in which the crop is exposed to prolonged periods of submergence and consumes
large amounts of water (Singh, 2013). In Eastern Asia, strategies that emphasize the use of younger
seedlings (<15 days) planted singly and at wider spacing, together with the adoption of intermittent
irrigation, organic fertilization, and active soil aeration have been proposed (Chapagain and Yamaji,
2010).
The objective of this study is to explore the impact of climate change on the WF of paddy rice
in Korea by using GCM ensemble climate projections.
2. Methods
2-1. Study area

Korea lies in the Far East and covers an area of about 45% of the Korean peninsula (98,477
km2), with altitude between 0 m and 1,950 m above sea level. Korea has eight administrative
provinces which can be studied independently, namely Gyeonggi (A), Gangwon (B), Chungnam (C),
Chungbuk (D), Gyeongbuk (E), Jeonbuk (F), Jeonam (G) and Gyeongnam (H) as shown in Fig. 1.
The climate of Korea is the Asian monsoon with average temperatures that range from about -10°C in
winter to about 30°C in summer. Two-thirds of the annual rainfall falls in the summer months, June to
September. Rice is the main staple crop in Korea. Approximately 80% of the paddy fields in Korea are
irrigated (Yoo et al., 2013b). The rice growing period is from May to September and monoculture is
the general practice. May is the nursery period and seedlings are transplanted in late May. About 77%
of the paddy fields lie on inceptisol soils. Irrigation water is supplied from May to September to keep
the water depth at 5–10 cm and rice is harvested in late October (Chung, 2013).
Legend
0 100 200 km Province
boundary
Major
paddy area
Fig. 1 Rice growing areas of Korea (Chung, 2013).
2-2. Climate data

Climate data (i.e. temperature, rainfall, radiation, wind speed and relative humidity) for eight
administrative provinces for 1971–2000 (baseline) were downloaded from the Korean Meteorological
Administration (www.kma.go.kr). Thereafter, climate data from 12 General circulation models
(GCMs) and 3 Representative Concentration Pathways (RCPs) for 1971–2000 (1985s), 2021–2040
(2030s), 2051–2070 (2060s) and 2080–2100 (2090s) were extracted from an online database courtesy
of the World Climate Research Program’s Working Group on Coupled Modeling (CMPI5). The
selected models are shown in Table 1. The mean of the absolute (for temperature) and relative
differences (rest of parameters) between the GCM baseline and future scenarios were used to perturb
the observed baseline using the Long Ashton Research Station weather generator (LARS-WG)..
Scenario refers to a single realization of the climate, generated for either the baseline or future climate
using global climate models (GCMs) to be input into the yield or net irrigation requirement (NIR)
model (Ledbetter et al., 2012). The LARS-WG is widely used for climate changes studies to generate
daily site-specific scenarios of future and past climate, by considering changes in both mean climate
and climate variability (Kim et al., 2013). Of the 3 RCPs used in this study, RCP2.6, RCP4.5 and
RCP8.5 represent the low, medium and high scenarios of radiative forcing and greenhouse gas
emissions examined by the climate modeling community. The relative changes in rainfall and absolute
changes in temperature were presented and analyzed in box and whisker plots.
Table 1 Selected GCMs

Model Modeling Center
CSIRO (Commonwealth Scientific and Industrial Research Organisation,
ACCESS1.0
Australia), and BOM (Bureau of Meteorology, Australia)
BCC-CSM1 Beijing Climate Center, China Meteorological Administration
CCSM4 National Center for Atmospheric Research
Centre National de Recherches Meteorologiques / Centre Europeen de Recherche

CNRM-CM5
et Formation Avancees en Calcul Scientifique
FGOALS-G2 LASG, Institute of Atmospheric Physics, Chinese Academy of Sciences
GFDL-CM3 Geophysical Fluid Dynamics Laboratory
GISS-E2 NASA Goddard Institute for Space Studies
Met Office Hadley Centre , Instituto Nacional de Pesquisas Espaciais, Korea
HADGEM-ES
Meteorological Administration
INMCM4 Institute for Numerical Mathematics
IPSL-CM5A Institut Pierre-Simon Laplace
Atmosphere and Ocean Research Institute (The University of Tokyo), National
MIROC5 Institute for Environmental Studies, and Japan Agency for Marine-Earth Science
and Technology
MRI-CGCM3 Meteorological Research Institute
2-3. Crop model: WARM

The Water Accounting Rice Model (WARM) is a simplified daily time step model for the
simulation of growth and development of paddy rice crops (Confalonieri, 2010). WARM has an
advantage over other crop models in that it has a user friendly interface and it is designed to account
solely for the micrometeorological peculiarities of paddy fields . WARM can also account for diseases
and cold shock induced spikelet sterility that in fluence the final yield (Confalonieri, 2009). Crop
development is based on the thermal time accumulated between a base temperature and a cutoff
temperature, optionally modulated by a photoperiodic factor. In crop growth, aboveground biomass
rate is calculated on a daily time step as shown below (eq. 1.):
=
AGBd RUEact × 0.5 ⋅ Rad ⋅ (1 − e − kLAI ) (1)
-2 -1 -kLAI
where Rad (MJ m d ) is daily global solar radiation, 1-e is the fraction of photosynthetically
active radiation (PAR) intercepted by the canopy, k is light extinction coefficient, LAI (m2 m-2) is the
leaf area index, RUEact (g MJ-1) is actual PAR use efficiency calculated as shown below (eq. 2):
RUE=
act RUEmax ⋅ Tlim ⋅ ( Rad _ F ) ⋅ ( DVS _ F ) (2)
-1
where RUEmax (kg MJ ) is the radiation use efficiency value (crop parameter) not limited by water,
nutrients, pest, diseases, senescence, excess of radiation, temperature, damages. Tlim, Rad_F and
DVS_F are unit less factors in the range 0 (maximum stress) – 1 (no limitation) accounting for
temperature limitations, saturation of the enzymatic chains, and senescence phenomena, respectively.
Tlim is temperature-limitation factor, calculated by the equation below (eq. 3);
C
 Tc −Tb

  T − Tb   Tc − T  To −Tb

= 
 To − Tb   Tc − Tc 
Tlim (3)

 
where T (ºC) is the mean daily air temperature, Tc (ºC) is the ceiling air temperature at which crop
growth ceases, To (ºC) is the optimal air temperature at which the maximum rate of development
occurs, Tb (ºC) is the base air temperature for crop growth, C is a shape factor set equal to 1.8 to allow
the beta-function to approximately cross linear progression between Tb and To at Tlim=0.5.
The factors accounting for saturation of the enzymatic chains involved with photosynthesis
(Rad_F) and for the effect of senescence (DVS_F) are calculated using empirical functions. Daily
aboveground biomass (kg m-2 d-1) is partitioned to leaves, panicles and stems. Partitioning to panicles
starts at the panicle initiation stage (PI) and peaks at the beginning of the ripening phase. A daily
factor accounting for spikelet sterility due to cold shocks during the period between panicle initiation
and heading is also calculated. The parameters used in this study are shown in Table 1. The parameters
in Table 2 were determined for the Korean Japonica rice from calibration, literature and model default
parameters (Confalonieri et al., 2006). For the parameter values in Table 2, the R2 ranged for rice
yields from 0.63 to 0.92 when simulated and observed values were compared. The authors assumed
that here is no significant variation in soil types, elevation and physical geomorphology across the 8
provinces.
Table 2 WARM parameters
Development
Value unit Growth parameters Value Unit
parameters
Base T before
12 °C Radiation use efficiency 2.4 g MJ-1
emergence
Max. T before Extinction coefficient for solar
40 °C 0.5 -
emergence radiation
GDDs to reach
80 °C day Base T for biomass accumulation 12 °C
emergence
Base T after Optimum T for biomass
12 °C 28 °C
emergence accumulation
Max T after Max T for biomass accumulation 35 °C
40 °C
emergence Leaf area index at emergence 0.02 m2 m-2
GDDs emergence – Specific leaf area at emergence 27 m2 kg-1
970 °C day
flowering Specific leaf area end tillering 18 m2 kg-1
GDDs flowering – Fraction of biomass partitioned to
550 °C day 0.8 -
maturity leaves at emergence
Development stage
2.0 - Kc full canopy 1.2 -
at harvest
2-4. Water footprint calculation

The procedure for the estimation of the WF of paddy rice was adopted from the WF
assessment manual (Hoekstra et al., 2011). According to the spatio-temporal explication in the manual,
this study is of Level B i.e. national scale. The study was limited to the direct consumption WF (blue
and green) and the the gray WF was neglected (Bocchiola et al., 2013). The green and blue WF
(WFgreen and WFblue, respectively) are given below (eq. 4 and eq. 5) (Hoekstra et al., 2011):
t
∑ min(WD, R eff )i
WFgreen = i =1
(4)
Y
n
∑ max(0,WD − R eff )i
WFblue = i =1
(5)
Y
where WD is the water demand and Reff is the effective rainfall for a daily time step i , n is the total
number of time steps and Y is the rice yield.
The water demand consists of (1) a land preparation demand of 140mm, (2) nursery or seedbed
demand and (3) a water requirement to maintain an inundated soil environment in a paddy field after
transplanting (Guerra et al., 1998). The water requirement after transplanting can be estimated by
multiplying the reference evapotranspiration (ETo) and the crop coefficient (Kc). ETo is based on the
atmospheric water demand and the crop and soil surface characteristics and was estimated using a
modified Microsoft Excel spreadsheet for computing the reference evapotranspiration (ETo) by using
the FAO Penman-Monteith equation (Lupia, 2013). Kc refers to the ratio of the actual ET to the ETo
and it is determined by the climate, crop’s conditions and the cultivation practices (Djaman and Irmak,
2013). The difference between the water demand and effective rainfall represents the net irrigation
requirement (NIR). Effective rainfall was estimated using a method for non-submerged rice based on
daily readings (Dastane, 1978). In this method, 80% of the daily rainfall is considered effective if it is
more than half of the daily reference evapotranspiration (ETo). The analysis of the WF was truncated
and transport, labor etc. were not considered to avoid double counting and for consistency with current
studies (Hoekstra et al., 2011; Sun et al., 2013a; Yoo et al., 2013a; Chapagain and Hoekstra, 2010).
2-5. Sensitivity analysis

To illustrate the possible changes in the rice yield and NIR, their sensitivity to a combination
of changes in rainfall and temperature were computed and illustrated by response surface diagrams.
Values of the baseline rainfall were multiplied by factors of 0.5 to 1.5 in increments of 0.25 while –
2°C to 7°C were added to the baseline temperature in increments of 0.5 °C.
2-6. Trend and uncertainty analysis

The impact of climate change on the NIR, yield, WF and WFblue/WFgreen ratio was then
assessed quantitatively and presented in box and whisker plots. To facilitate for easier comparison
across time periods and provinces, the results of the different scenarios will be presented as relative
(future/baseline) or absolute changes (future – baseline; for temperature only). In the box and whisker
plots in this study, the box plot whiskers indicate the full data range, the box shows the interquartile-
range and the line across the boxes represents the mean. The diagram for the WFblue/WFgreen however,
the length box is twice the standards deviation for the time series for each scenario. For each scenario,
the means were used to detect trend patterns while the ranges and interquartile ranges or standard
deviation were used to assess for the GCM ensemble uncertainty. Uncertainty refers to any departure
from the unachievable ideal of completely deterministic knowledge of a system (Asseng et al., 2013).
3-1. Future climate

Fig. 2 shows the projected changes in temperature and rainfall. The 12 GCMs indicate
significant warming of the country in the future particularly in the 2090s. The RCP2.6 peaks in the
2060s and the increasing trend slightly declines in the 2090s. The trends in the temperature follow the
trends in the radiative forcing. The model disagreement (uncertainty) also increases in the future and is
highest in the 2090s. Temperature uncertainty ranges from 0.6 to 1.6°C. Generally, rainfall was
projected to increase in the future. There are strong increasing trends that persist until the 2060s and
decline thereafter towards the 2090s. The uncertainty troughs in the 2060s and is highest in the 2090s.
Rainfall uncertainty ranges from 0.06 to 0.13.
Fig. 2 Projected changes in temperature (°C) and rainfall
3-2. Crop model sensitivity analysis

The response surface diagrams presented in Fig. 3 demonstrate the sensitivity of the paddy
rice net irrigation water requirement (mm) and potential yield (t ha-1) to changes in ambient air
temperature and rainfall. The potential yield is shown to peak at an increase in temperature of 1°C and
decreases as temperature increases further. This is because rice matures quicker at higher temperatures,
which shortens the duration of key growth phases resulting in reduced grain yield (Semenov et al.,
2012). The NIR increases as the temperature increases due to the increased evaporative demand. On
the other hand, the NIR decreases as the rainfall increases.
Fig. 3 Yield and Net Irrigation Requirement (NIR) responses to changes in temperature and
rainfall
3-3. Paddy rice yield

Fig. 4 shows the variation of the simulated changes in the long-term yield for the 2030s, 2060s
and 2090s compared to the baseline (1971 – 2000). Yield for the RCP2.6 generally remained more or
less equal to the baseline values. There is convergence for the RCP4.5 and RCP8.5 that the paddy rice
yields will decrease in the future. The decreases in the yield are higher in the RCP8.5 than for the
RCP4.5. The changes in yield ranged from 0.93 – 1.08, 0.87 – 1.07 and 0.69 – 1.07 for RCP2.6,
RCP4.5 and RCP8.5, respectively. Similarly, the variability (uncertainty) in the simulated yield
increases in the future and with the RCPs. The interquartile range ranged from 0.02 – 0.07, 0.01 – 0.12
and 0.01 – 0.23 for RCP2.6, RCP4.5 and RCP8.5, respectively. Although the differences across the
provinces appear small in relative values presented in Fig. 4, the absolute difference of yields between
the provinces is large. Overall, the rice yields in Jeonbuk will be most affected while the yield
Gyeongnam and Chungnam will be least affected.
Fig. 4 Projected changes in the simulated paddy rice yield for (a) RCP8.5 (b) RCP4.5 (c) RCP2.6
3-4. Net irrigation water requirement
Fig. 5 shows the variation in the simulated changes in NIR. There is general model agreement
and the NIR is projected to increase in the future scenarios. There is a continuing increasing trend for
the RCP4.5 and RCP 8.5. For the RCP2.6, the NIR increases from the baseline values, peaks in the
2060s and slightly declines towards the 2090s. The changes in the NIR ranged from 1.0 – 1.27, 1.02 –
1.24 and 0.99 – 1.26 for the RCP2.6, RCP4.5 and RCP8.5, respectively. The trend patterns shown by
the NIR can be attributed to the increases in temperature and subsequently evapotranspiration demand
for the different RCPs. The interquartile range ranged from 0.04 – 0.20, 0.04 – 0.22 and 0.02 – 0.21
for the RCP2.6, RCP4.5 and RCP8.5, respectively. There is significantly higher changes and
uncertainty (range and interquartile range) for the NIR than for the yield.
Fig. 5 Projected changes in the simulated Net Irrigation Requirement (NIR) for (a) RCP8.5 (b)
RCP4.5 (c) RCP2.6
3-5. Water footprint

Fig. 6 shows the variation in the simulated changes in the water footprint (WF). The WF was
projected to roughly remain at current levels for all time periods in the RCP2.6 with little uncertainty.
For the RCP4.5 and RCP8.5 however, the WF was projected to decrease in the 2030s and increase in
the 2060s and is highest for the 2090s. Overall, the uncertainty increases with each time step and RCP
as shown in Fig. 6. With the exception of the north eastern provinces of Chungbuk, Gangwon and
Gyeongbuk, the WF in the 2030s is lower than that of the baseline. As shown in the sensitivity
analysis, small changes in temperature (typical of the 2030s) will increase yields and therefore reduce
the WF.
Fig. 6 Projected changes in the simulated water footprint (WF) for (a) RCP8.5 (b) RCP4.5 (c)
RCP2.6
3-6. WFblue/WFgreen ratio

Finally, Fig. 7 shows the variation in the WFblue/WFgreen ratio. The ratio correlates with aridity
(Siebert and Döll, 2010) with low values in the Gyeonggi and Jeonbuk provinces while higher values
are in the Gyeongbuk and Chungnam provinces. Overall, for the RCP2.6, the WFblue/WFgreen ratio will
increase and peak in the 2060s and decline thereafter in the 2090s. The ratio will not change
significantly for the RCP4.5 but will steadily increase in the future for the RCP8.5. Therefore, climate
change will not only increase the WF but also modify the ratio of WFblue to WFgreen which are inversely
correlated (Marta et al., 2012). The Chungbuk, Chungnam and Gyeongbuk province were vulnerable
to high WFblue/WFgreen ratios above 1. These ratios would imply that more water is required from
irrigation than rainfall. Such a scenario would nullify some advantages of growing crops in the high
ET rainy season from a water resource standpoint alone. Among the RCPs used in this study, the
RCP2.6 gives the lowest WF and the best WFblue/WFgreen ratio while RCP8.5 gives the worst. A lower
ratio is desirable to minimize environmental impacts and withdrawal and conveyance costs of blue
water. Climate change mitigation therefore is required to avoid the potential impacts of high forcing
scenarios such as RCP8.5. Further research to quantify the amount of blue and green water used in
agriculture and adjust agricultural planting to increase the green water proportion in crop production
should be carried out (Sun et al., 2013b)
Fig. 7 Projected changes in the WFblue/WFgreen ratio for (a) RCP8.5 (b) RCP4.5 (c) RCP2.6
In this paper, the direct effect of CO2 and the advances in technology on future rice yields and
water application were not considered. The neglect of direct CO2 fertilization is justified because
research has shown that although paddy rice yield increases significantly by doubling the CO2
concentration at current temperatures, if the temperature also increases, there is no CO2 enrichment
effect on rice yield at higher temperatures (Kim et al., 1996; Ziska et al., 1997; Yang et al., 2013).
However advances in technology irrigation technology will improve the water use efficiency.
4. Conclusion
This paper outlines techniques for generating ensembles of climate change scenarios and
projecting the potential impacts of climate on the yield, irrigation water demand and ultimately the
water footprint. There was general agreement from the 12 GCM ensemble that both temperature and
rainfall will increase in the future. Simulations with the Water Accounting Rice Model (WARM) and
water balance model showed that the increased temperatures would lead to reduced rice yield and
higher irrigation demand. In addition to the water footprint being projected to increase, the blue to
green water footprint component ratio was also projected to be altered in the future. Climate change
potentially threatens rice production in the Chungbuk, Chungnam and Gyeongbuk provinces. Further
research should be carried out to investigate how Korean farmers can reduce blue water withdrawal
and increase green water use in the future.
References
Angulo, C., Rötter, R., Lock, R., Enders, A., Fronzek, S., and Ewert, F., 2013, “Implication of crop
model calibration strategies for assessing regional impacts of climate change in Europe.”
Agricultural and Forest Meteorology, 170, 32-46.
Asseng, S., Ewert, F., Rosenzweig, C., Jones, J. W., Hatfield, J. L., Ruane, A. C., ... and Williams, J.
R., 2013, “Uncertainty in simulating wheat yields under climate change.” Nature Climate Change,
3, 827–832.
Bocchiola, D., Nana, E., and Soncini, A., 2013, “Impact of climate change scenarios on crop yield and
water footprint of maize in the Po valley of Italy.” Agricultural Water Management, 116, 50-61.
Bouwer, H., 2000, “Integrated water management: emerging issues and challenges.” Agricultural
Water Management, 45(3), 217-228.
Chapagain, T., and Yamaji E., 2010, “The effects of irrigation method, age of seedling and spacing on
crop performance, productivity and water-wise rice production in Japan.” Paddy Water
Environment, 8, 81–90.
Chung, S.-O., 2013, “Projecting Future Paddy Irrigation Demands in Korea.” Irrigation and Drainage,
62 (3), 297–305.
Confalonieri R., Gusberti D., and Acutis M., 2006, “Comparison of WOFOST, CropSyst and WARM
for simulating rice growth (Japonica type – short cycle varieties).” Italian Journal of
Agrometeorology, 11-3, 7-16.
Confalonieri, R., Acutis, M., Bellocchi, G., and Donatelli, M., 2009, “Multi-metric evaluation of the
models WARM, CropSyst, and WOFOST for rice.” Ecological Modelling, 220(11), 1395-1410.
Confalonieri, R., Bellocchi, G., Bregaglio, S., Donatelli, M., and Acutis, M., 2010, “Comparison of
sensitivity analysis techniques: a case study with the rice model WARM.” Ecological Modelling,
221(16), 1897-1906.
Dastane, N.G., 1978, “Effective Rainfall in Irrigated Agriculture. FAO Irrigation and Drainage Paper
No. 25.” Food and Agriculture Organization (FAO), Rome, Italy, 67pp.
Djaman, K. and Irmak, S., 2013, “Actual Crop Evapotranspiration and Alfalfa- and Grass-Reference
Crop Coefficients of Maize under Full and Limited Irrigation and Rainfed Conditions.” Journal
of Irrigation and Drainage Engineering, 139(6), 433–446.
Guerra, L., Bhuiyan, S.I. Tuong, T.P. and Barker, R., 1998, Producing more rice with less water.
SWIM Paper 5. Colombo, Sri Lank: International Water Management Institute, 4 – 6.
Hoekstra, A.Y., 2003, “Virtual water trade.” In: Proceedings of the international expert meeting on
virtual water trade, Delft, the Netherlands, 12-13 December 2002. Value of research report
series No. 12, UNESCO-IHE, Delft, The Netherlands.
Hoekstra, A.Y., Chapagain, A.K., Aldaya, M.M., Mekonnen, M.M., 2011, The water footprint
assessment manual: setting the global standard. Earthscan publishing, Washington, USA.
Kim, H.K., Parajuli, P.B., and Filip To, S.D., 2013, “Assessing impacts of bioenergy crops and climate
change on hydrometeorology in the Yazoo River Basin, Mississippi.” Agricultural and Forest
Meteorology, 169, 61-73.
Kim, H.Y., Horie, T., Nakagawa, H., and Wada, K., 1996, “Effects of elevated CO2 concentration and
high temperature on growth and yield of rice [Oryza sativa], 2: The effect on yield and its
components of Akihikari rice.” Japanese Journal of Crop Science,65(4),644-651.
Kim, J., Shon, J., Lee, C.K., Yang, W., Yoon, Y., Yang, W.H., and Lee, B.W., 2011, “Relationship
between grain filling duration and leaf senescence of temperate rice under high temperature.”
Field Crops Research, 122(3), 207-213.
Ledbetter R., Prudhomme, C., and Arnell N., 2012, “A method for incorporating climate variability in
climate change impact assessments: Sensitivity of river flows in the Eden catchment to
precipitation scenarios.” Climatic Change, 113(3-4), 803-823.
Lupia, F., 2013, ETo-PM version 0.9. http://dspace.inea.it, July, 2013.
Marta, A.D., Mancini, M., Natali, F., Orlando, F., and Orlandini, S., 2012, “From water to bioethanol:
The impact of climate variability on the water foot print.” Journal of hydrology, 444-445,180-186.
Mehta, V.K., Haden, V.R., Joyce, B.A., Purkey, D.R., and Jackson, L.E., 2013, “Irrigation demand
and supply, given projections of climate and land-use change, in Yolo County, California.”
Agricultural Water Management, 117, 70-82.
Rockström, J., Folke, C., Gordon, L., Hatibu, N., Jewitt, G., Penning de Vries, F., and Schulze, R.,
2004, “A watershed approach to upgrade rainfed agriculture in water scarce regions through
Water System Innovations: an integrated research initiative on water for food and rural
livelihoods in balance with ecosystem functions.” Physics and Chemistry of the Earth, Parts
A/B/C, 29(15), 1109-1118.
Semenov, M.A., Mitchell, R.A.C., Whitmore, A.P., Hawkesford, M.J., Parry M.A.J., and Shewry,
P.R. , 2012, “Shortcomings in wheat yield predictions.” Nature Climate Change, 2(6), 380-382.
Siebert, S., and Döll, P., 2010, “Quantifying blue and green virtual water contents in global crop
production as well as potential production losses without irrigation.” Journal of Hydrology, 384,
198–217.
Singh, Y.V., 2013, “Crop and water productivity as influenced by rice cultivation methods under
organic and inorganic sources of nutrient supply.” Paddy and Water Environment, 11,531–542.
Sun, S., Wu, P., Wang, Y., Zhao, X., Liu, J., and Zhang, X., 2013a, “The impacts of interannual
climate variability and agricultural inputs on water footprint of crop production in an irrigation
district of China.” Science of the Total Environment, 444, 498-507.
Sun, S.K., Wu, P.T., Wang, Y.B., and Zhao, X.N., 2013b, “Temporal Variability of Water Footprint
for Maize Production: The Case of Beijing from 1978 to 2008.” Water Resources Management,
1-17.
Supit, I., Van Diepen, C.A., De Wit, A.J.W., Wolf, J., Kabat, P., Baruth, B., and Ludwig, F., 2012,
“Assessing climate change effects on European crop yields using the Crop Growth Monitoring
System and a weather generator.” Agricultural and Forest Meteorology, 164, 96-111.
Tallec, T., Béziat, P., Jarosz, N., Rivalland, V., and Ceschia, E., 2013, “Crops’ water use efficiencies
in temperate climate: Comparison of stand, ecosystem and agronomical approaches.” Agricultural
and Forest Meteorology, 168, 69-81.
Van Minnen, J.G., Alcamo, J., and Haupt, W., 2000, “Deriving and applying response surface
diagrams for evaluating climate change impacts on crop production.” Climatic Change, 46, 317–
338.
Yang, W., Cho, H.S., Kim, M., Seong, K.Y., Park, T.S., Seo, M.C., and Kang, H.W., 2013, “Re-
examination of the standard cultivation practices of rice in response to climate change in Korea.”
Journal of Crop Science and Biotechnology, 16(2), 85-92.
Yoo, S. H., Choi, J.Y., Lee, S.H., and Kim, T., 2013a, “Estimating water footprint of paddy rice in
Korea.” Paddy and Water Environment, 1-12.
Yoo, S.-H., Choi J.-Y., Lee S.-H., Oh Y.-G., and Yun D.-K., 2013b, “Climate change impacts on
water storage requirements of an agricultural reservoir considering changes in land use and rice
growing season in Korea.” Agricultural Water Management, 117, 43-54.
Ziska, L.H., Namuco, O., Moya, T., and Quilang, J., 1997, “Growth and Yield Response of Field-
Grown Tropical Rice to Increasing Carbon Dioxide and Air Temperature.” Agronomy Journal,89,
45–53.
THE FUTURE
F-03
Effect of Return Flow on Water Temperature in

Irrigation-drainage Canal under Spill-over Paddy Irrigation
Masaomi Kimura*, Kouki Kasai**, Toshiaki Iida*, Marie Mitsuyasu* and

Naritaka Kubo
*Graduate School of Agricultural and Life Sciences, the University of Tokyo 1-1-1
Yayoi, Bunkyo-ku, Tokyo 113-8657, JAPAN
**Faculty of Agriculture, the University of Tokyo 1-1-1- Yayoi, Bunkyo-ku, Tokyo
113-8657, JAPAN
ABSTRACT
High temperature damage to rice reduces farmers’ income by degrading rice quality and it may be
magnified by climate change in the future. Among several proposed measures against it, paddy water
management technique like spill-over irrigation is one of the simplest ways for famers. Fresh
irrigation water continues to be taken to paddy fields under spill-over irrigation on hot days after rice
heading. Hence the method requires a large quantity of water which is sufficiently cool for high
efficiency. However it is considered that in summer paddy fields in downstream area can only take
higher temperature irrigation water than in upstream especially in dual-purpose canal system, which
is used as both irrigation and drainage canal.In this study, water temperature variation along open
channel system was observed in Tedori-gawa Shichika irrigation area, which is located in Ishikawa
prefecture, Japan. In the irrigation area dual-purpose canal system is used and spill-over irrigation is
suggested on hot days. As a result of field observation, it has become clear that there is a large
increase in water temperature relative to flow distance especially in downstream part. The water
temperature increase is thought to be due to both climate factors and return flow from paddy fields. In
order to analyze this phenomenon more precisely and forecast the situation in the future, a water
temperature simulation model was proposed which is based on heat balance in open channel.
Accordingly, the extent of influence of return flow on water temperature in dual-purpose canal was
estimated.
Keywords: Irrigation water temperature, dual-purpose canal, spill-over irrigation, high temperature
damage to rice, Tedori-gawa Shichika irrigation area
1. Introduction
Recently, high temperature damage to rice grain ripening has been a big concern in Japan. It
raises yield reduction, lack of grain filling, and occurrence of cracked rice and milky white kernel,
which results in degrading rice classification. Extremely hot summer in 2010 brought serious decrease
in first-class rice rate down to 62%, which is ordinarily about 80% (Nakagawa, 2013). High
temperature damage is thought to be mainly caused by high temperature on days after rice heading
(Morita, 2008) and milky white kernel partly looks chalky because failure of growth produces voids
inside amyloplast of rice endosperm (Tashiro and Ebata, 1975). Nowadays, some effective measures to
prevent the damage are proposed; cultivar improvement, fertilizer management, delay of rice planting,
paddy water management, and so on. Among them, the measure by paddy water management is indeed
a one of simplest ways for famers. Especially, spill-over irrigation, in which irrigation to paddy fields
and over-flow drainage are continuous over a period of time, is known as a very effective method and
attracting much attention. Nagahata et al. (2005) indicated that spill-over irrigation management
during 30 nights (from 6 pm to 6 am) after rice heading obviously reduced the occurrence of cracked
rice and milky white kernel compared to ordinary paddy water management.
However, the method requires a large amount of irrigation water which is sufficiently cool.
Considering irrigation facility’s capacity and water rights, it is doubtless that not every farmer who
wants to adopt spill-over irrigation is always able to practice the method. Moreover, much attention
should be paid to the mechanism of water temperature formation in irrigation canal since irrigation
water temperature is a significant factor for determining the effect of spill-over irrigation against high
temperature damage to rice. It is presumed that water temperature in open channel is affected by
climatic condition and becomes higher towards downstream in summer. In addition, return flow, that is,
drainage water from paddy fields in upstream area flowing into irrigation canal in downstream,
probably influences the water temperature in downstream. Hence, especially in dual-purpose canal
system, which is used as both irrigation and drainage canal, it is considered that in summer paddy
fields in downstream area can only take higher temperature irrigation water than in upstream.
In this study, water temperature variation in dual-purpose canal was precisely observed along a
main canal of Tedori-gawa Shichika irrigation area in Japan. The water level and temperature variation
of branch drainage channel was also investigated so that the phenomena of increase in irrigation water
temperature by return flow were certainly grasped. Then a water temperature simulation model in
dual-purpose canal was proposed which is based on heat advection with water flow and heat balance
in open channel. As a result of calculation by the model, the amount of return flow from paddy fields
was estimated. Then, the simulation model was improved by incorporating the effect of return flow
and the verification of the model was done by comparing calculation results with observational data.
2. Methodology of observation
2.1 Investigation site

Tedori-gawa Shichika irrigation area is located in the alluvial fan of downstream of Tedori river
in Ishikawa prefecture, Japan. Irrigation water is taken from Sirayama head works at the top of the
alluvial fan and delivered to about 5,000 ha of paddy fields through branched canals. In the irrigation
area, dual-purpose canal system is used and irrigation-drainage channels are laid throughout whole the
area. Usually, rice seedlings are planted in early May, mid-summer drainage practice is done in late
June, ears of rice appear in late July and harvesting season starts in late August. In order to prevent the
high temperature damage to rice, no-flooding water management is encouraged after mid-summer
drainage and spill-over irrigation is recommended at nights on especially hot days.
Fig. 1 Map of the investigation site
Note) Black solid line represents the observational main canal, points A-G, 1-6 and I-IV indicate the observation plots in the
main canal, on the paddy fields, and the division works, respectively, and points U, R and W are the observation plot at
downstream of the head works, the confluence point of return flow and main flow, and the place of the weather station
(Ishikawa Prefectural University), respectively.
2.2 Observation in a main irrigation-drainage canal

The observation plots A-G along a main canal are also shown in Fig. 1. The observational
irrigation-drainage canal is one of main canals of the irrigation area which is numbered 4-1, 4-2. The
irrigation water of 367.7 ha of paddy fields is taken from the canal and the drainage water from 246.8
ha flows into the canal. At the observation plots A-G along the main canal and plot U at downstream
of the head works, a pressure type water level and temperature data logger (HOBO U20-001-04;
Onset) was set inside a polyvinyl chloride pipe fixed on side wall of the canal. The water level and
water temperature data at 30 min intervals were obtained from April 2011 to September 2012. All the
plots A-G were located at just upstream of drops so that the discharge of flow can be converted from
the water depth by using calculation of non-uniform flow between the place of the logger and the drop
or formula of weir.
The water level data at the division works I-IV in Fig. 1 at 1 hr intervals were obtained from
Tedori-gawa Shichika land improvement district. Meteorological data (solar radiation, air temperature,
atmospheric pressure, wind speed, and relative humidity) were measured by a weather station set at
Ishikawa Prefectural University (point W in Fig. 1).
2.3 Observation in a branch drainage channel

In order to catch the confluence phenomena of return flow, an auto running camera (Garden
Watch Cam; Brinno) aiming at a scale fixed on side wall of an observational drainage channel and a
water temperature logger (TR-52S; T&D) were set at the plot R in Fig. 1. Fig. 2 shows the pictures of
the observation system at the plot R. Only the drainage water from about 4 ha of paddy fields (no
excess irrigation water) flows in the drainage channel and drops into the main canal at the plot R. The
same type of water temperature logger as the plot R was also set at division works IV in order to
obtain the water temperature in main canal before the return flow joins at the plot R. There wasn’t any
other confluence point but the plot R between the points IV and E.
Fig. 2 Pictures of the observation system at the plot R
Note) Gray arrow and white arrow indicates the flow of the main canal and drainage water, respectively.
2.4 Observation on paddy plots

The flood water depth and temperature were observed on the paddy fields 1-6 in Fig. 1 so that
the characteristics of the temperature of surface drainage water were Fig.d out. A pressure type water
gauge (HTV-020KP; Sensez) and a temperature logger (Thermochron SL type; KN Laboratories) were
set at the outlet of each paddy plots. The temperature logger was fixed with a floating board made with
foamed polystyrene so that it could measure the surface water temperature of the flood water. The
flood water depth and temperature data at 1 hr intervals were obtained during irrigation period in
2012.
3.1 Water temperature variation along the main irrigation-drainage canal

Fig. 3 (a) and (b) show the daily averaged water temperature at the plots A-G, U and air
temperature at the point W from July to August 2012. The observed water temperature in the main
canal got higher relative to the plot U toward downstream and generally went up and down along with
air temperature. Fig. 3 (c) shows the daily averaged discharge at the division works I. The water
temperature in the main canal tended to increase when the flow rate was low, which means when the
intake from the head works was stopped or regulated because of large runoff from the basin of Tedori
river.
U A B C D E F G
26
Water Temperature (deg C)
24
22
20
18
16
14
1-Jul 11-Jul 21-Jul 31-Jul 10-Aug 20-Aug 30-Aug
32
(deg C)
28
24
Air temperature
20
8
6
(m3/s)
4
2 Discharge at I
0
Fig. 3 Daily averaged observational results, July - August 2012 ((a): Water temperature, (b): Air
temperature, (c): Discharge at I)
The water temperature variations at each observation plot from 30 July to 5 August 2012 are
shown in Fig. 4. The difference of water temperatures between the plots U and G varied from about 2
deg C at nights up to about 5 to 6 deg C around noon. Especially in downstream part, there was a large
increase in water temperature relative to flow distance along the main canal. This phenomenon is
thought to be explained by the characteristics of open channel for irrigation. Generally in irrigation
canal, water depth gets small and flow speed becomes slow as it goes downstream. Therefore the
water temperature in downstream part is more sensitive to meteorological conditions and return flow
from paddy fields than upstream.
In order to Fig. out the heat balance of water in the main canal consist of the factors other than
return flow, the terms of the following equation were calculated. Given that there is no return flow into
the main canal, the heat balance equation of water in the canal is described with Lagrangian method as
Rn − H − LE − G = W (1)
where Rn is the net radiation (W/m2), H is the sensible heat flux (W/m2), LE is the latent heat flux
U A B C D E F G
28
Water Temperature (deg C)
26
24
22
20
18
30-Jul 31-Jul 1-Aug 2-Aug 3-Aug 4-Aug 5-Aug 6-Aug
0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00
Fig. 4 Water temperature variations at the observation plots, 30 July - 5 August 2012
(W/m2), G is the heat flux into the ground (W/m2), and W means the amount of the heat storage change
of the water in the main canal per unit time (W/m2). Rn was calculated by using the equation:
R n = (1 − α )S + Ld − εσ Tw
4
(2)
where α is the albedo of water surface (0.06), S is the solar radiation (W/m2), Ld is the downward
long-wave radiation (W/m2), ε is the emissivity of water (0.96), σ is the constant of Stefan-Boltzmann
(5.67 × 10-8) (W/(m2·K4)), and Tw is the water temperature (K). H and LE were calculated by using the
bulk transfer equations and Ld was calculated by reference to Kondo (1995). G was calculated
assuming that the underground temperature below the channel bed had vertically linear distribution.
Fig. 5 shows the calculated variation of each term in (eq. 1) by using the water temperature at
the plot G from 30 July to 2 August 2012. The amount of the sensible heat H and the latent heat LE
were almost in counterpoise and they had low impact to the water temperature as with the heat flux
into the ground G. Moreover, it has become clear that the heat storage change of the water W varied
almost like the net radiation Rn and its value was close to zero at nights. Therefore, it is considered that
at least the increase of water temperature at nights shown in Fig. 4 was caused by the factors other
than the meteorological condition, that is, return flow from paddy fields.
Rn H LE G W
800
600
Heat flux (W/m2)
400
200
-200
30-Jul 0:00 31-Jul 0:00 1-Aug 0:00 2-Aug 0:00 3-Aug 0:00
Fig. 5 Calculated heat fluxes at the plot G, 30 July - 2 August 2012
3.2 Water level and temperature in the branch drainage channel
There were some events of confluence of much drainage water into the main canal at the point R
observed by the auto running camera on sunny days. Fig. 6 (a) and (b) show the water temperatures at
the plot IV, E and the depth of drainage channel made out by the images taken by the camera on 18
June and 5 July 2012. The water temperature at the plot IV means the water temperature before the
drainage water flows in at the point R and the water temperature at the plot E is that of after it flows in.
Obviously there was increase in water temperature between IV and E when and after the depth of the
drainage channel become large. The phenomena of water temperature increase in the main canal by
return flow from paddy fields were certainly Fig.d out by this observation.
The discharge of the drainage water at the confluence point R can be calculated by using the heat
conservation equation as follows:
QIV (TE − TIV )

Qin = (3)
Tin − TE
where Qin is the discharge of return flow, QIV is the discharge at the plot IV, Tin is the temperature of
drainage water, and TIV and TE is the water temperature at the plot IV and E, respectively. Fig. 7 (a)
and (b) show the estimated discharge of return flow at the confluence point R and the depth of the
drainage channel on 18 June and 5 July 2012. The both data have the similar type of variation.
Although it’s a comparison of discharge and depth of drainage water, it has become clear that the way
of estimation of return flow rate by using the heat conservation equation like (eq. 3) is effective to
some extent.
Fig. 8 (a) and (b) show the temperature of drainage water at the confluence point R and air
temperature from 10 to 19 June and from 15 to 25 July 2012. In Fig. 8 (a), there is large difference
between the temperature variation of air and of drainage water. However, in Fig. 8 (b), the both data
have similar variation. The drainage water temperature was close to the air temperature after mid-July.
IV IV
E E
Water temperature (deg C)
Depth of the drainage channel (right axis) Depth of the drainage channel (right axis)
17 80 20 80
16 60 19 60
Depth (cm)
Depth (cm)
15 40 18 40
14 20 17 20
13 0 16 0
0:00 6:00 12:00 18:00 0:00 0:00 6:00 12:00 18:00 0:00
Fig. 6 Water temperatures at the plot IV, E and depth of the drainage channel (right axis) ((a): 18 June
2012, (b): 5 July 2012)
Estimated discharge of return flow Estimated discharge of return flow
Depth of the drainage channel (right axis) Depth of the drainage channel (right axis)
0.20 80 0.20 80
0.15 60 0.15 60
Depth (cm)
Depth (cm)
(m3/s)
(m3/s)
0.10 40 0.10 40
0.05 20 0.05 20
0.00 0 0.00 0
0:00 6:00 12:00 18:00 0:00 0:00 6:00 12:00 18:00 0:00
Fig. 7 Estimated discharge of return flow at the confluence point R and depth of the drainage channel
(right axis) ((a): 18 June 2012, (b): 5 July 2012)
Drainage water temperature Air temperature

34
Temperature (deg C)
32
30
28
26
24
22
20
18
10-Jun 11-Jun 12-Jun 13-Jun 14-Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun
0:00 0:00 Drainage
0:00 water
0:00temperature
0:00 0:00 0:00 Air0:00 0:00
temperature 0:00
36
Temperature (deg C)
34
32
30
28
26
24
22
20
15-Jul 16-Jul 17-Jul 18-Jul 19-Jul 20-Jul 21-Jul 22-Jul 23-Jul 24-Jul 25-Jul
0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00
Fig. 8 Temperature of drainage water at the confluence point R and air temperature ((a): 10 – 19 June
2012, (b): 15 – 25 July 2012)
3.3 Water temperature at the outlet of the paddy plots

Fig. 9 (a) and (b) show the water temperatures at the outlet of the paddy plots 1-6 and the air
temperature from 22 to 29 June and from 30 July to 6 August 2012, respectively. Only the data of
when the flood water depth was larger than 1 cm are plotted in the graph so that the plotted data are
certainly of the flood water temperatures. The similar results as Fig. 8 were observed also by this
survey. Before mid-July, the amplitude of temperature variation of the flood water at the outlet was
obviously larger than that of air temperature. On the other hand, after mid-July, the both variation
comes closer. It is considered that this trend is because of the change of coverage by rice plant. As rice
plant grows, the solar radiation toward the surface of flood water and long-wave radiation are
intercepted by plant body. Then the temperature of flood water becomes more impervious against
external factors. The similar observed results were also reported by Iwakiri (1964), Oue and Kamii
(2002), and so on.
4. Simulation model of water temperature in the irrigation-drainage canal
4.1 Calculation of water temperature variation caused by climatic factor

As discussed in (§ 3.1), the increase in water temperature along the main canal is caused by both
meteorological condition and return flow from paddy fields. In order to analyze the amount of the
increase in water temperature by each factors quantitatively, the water temperature variation which is
caused by climatic factor was numerically estimated as follows.
Under the assumption that there is no return flow into the main canal, the equation of heat
conservation law of water which flows in open channel with Eulerian method is described as
∂Tw Q ∂Tw W
+ = (4)
∂t A ∂x ρ w cw D
Air temperature 1 2 3 4 5 6
38
Temperature (deg C)
34
30
26
22
18
14
22-Jun 23-Jun 24-Jun 25-Jun 26-Jun 27-Jun 28-Jun 29-Jun 30-Jun
0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00
38
Temperature (deg C)
34
30
26
22
18
14
30-Jul 31-Jul 1-Aug 2-Aug 3-Aug 4-Aug 5-Aug 6-Aug 7-Aug
0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00
Fig. 9 Water temperatures at the outlet of the paddy plots when the flood water depth was larger than 1
cm ((a): 22 – 29 June 2012, (b): 30 July – 6 August 2012)
where x indicates the flow distance along the main canal (m), Q is the discharge of main flow (m3/s), A
is the cross-sectional area of main flow (m2), ρw is the density of water (kg/m3), cw is the specific heat
of water (J/(kg·K)), and D is the hydraulic depth (m). Here the heat diffusion term is omitted since it’s
negligible compared to other terms. The water temperature variation at the plot G which is affected by
climatic factor was numerically obtained by discretizing (eq.4) by CIP method.
Fig. 10 (a) and (b) show the calculated water temperature variation at the plot G considering
only climatic factor from 30 July to 6 August 2012. The calculated data means the water temperature
at the plot G which is thought to be of the case if there is no return flow. The difference between the
observed temperature and the calculated temperature at the plot G means the estimated amount of the
increase in water temperature by return flow which flowed in the main canal between the plot A and G.
The amount of increase by return flow was about 2-3 dig C and it accounted for the main part of the
total increase in water temperature at nights.
A_obs G_obs G_cal (with climatic factor)

28
26
24
22
20
18
30-Jul 31-Jul 1-Aug 2-Aug 3-Aug 4-Aug 5-Aug 6-Aug 7-Aug
0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00
Fig. 10 Calculated water temperature variation at the plot G which is affected by climatic factor, 30 July –
6 August 2012
4.2 Estimation of return flow rate

The discharge rate of return flow which flows in the main canal was estimated by following
method. For instance between the plot F and G, the return flow rate QR (m3/s) can be calculated by the
equation of heat conservation:
TwG_obs − (TwF_obs + ∆TwFG_cal )

QR = QG_obs (5)
Tin − (TwF_obs + ∆TwFG_cal )
where “obs” indicates the observed value, ΔTwFG_cal is the calculated increase in water temperature
caused by climatic factor, and Tin is the temperature of return flow. The air temperature was applied to
Tin according to the discussion in (§ 3.3).
Fig. 11 shows the estimated return flow rates per paddy fields area which flowed in each interval
separated by the observation plots in downstream part of the main canal from 25 July to 5 August 2012.
The estimated rates of return flow were approximately same in each interval and its average amount
was calculated to be 1.55 × 10-4 mm/s which is equal to 13.4 mm/d during the term from 24 July to 7
August 2012. This value 13.4 mm/d is thought to represent the characteristic of the return flow in the
irrigation area in this term. Therefore it was named as the parameter of the return flow KR.
D - IV IV - E E-F F-G
paddy field area (mm/s)
8.0E-04
Return flow rate per
7.0E-04
6.0E-04
5.0E-04
4.0E-04
3.0E-04
2.0E-04
1.0E-04
0.0E+00
25-Jul 26-Jul 27-Jul 28-Jul 29-Jul 30-Jul 31-Jul 1-Aug 2-Aug 3-Aug 4-Aug 5-Aug 6-Aug
0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00
Fig. 11 Estimated return flow rate per paddy field area, 25 July – 5 August 2012
4.3 Simulation model incorporated the effect of return flow and its validation
The modified model of analyzing the water temperature variation in the dual-purpose canal
considering the effects of both climatic factor and return flow is described by using the parameter of
return flow KR as:
∂Tw Q ∂Tw W q
+ = + in (Tin − Tw ) (6)
∂t A ∂x ρ w cw D A
where qin (m2/s) is the discharge rate of return flow per unit length of channel which can be calculated
by the equation:
qin = K R × 10 −3 ÷ 86400 × Area ÷ L (7)
where Area is the area of paddy fields which discharge the return flow to the target interval of the main
canal (m2), and L is the length of the target interval of the main canal (m). Since the parameter fitting
term was 2012, the parameter validation term was selected to summer season in 2011.
Fig. 12 (a) and (b) show the calculated water temperature variation at the plot G considering
both climatic factor and return flow from 9 to 17 July and from 8 to 16 August 2011, respectively. The
calculated value by modified model (eq. 6) apparently represents the observed data at the plot G well
especially in mid-August 2011. The RMSE between the calculated and the observed water temperature
at the plot G was 0.80 deg C in mid-July and 0.51 deg C in mid-August.
A_obs
G_obs
G_cal (with climatic factor)
G cal (with climatic factor and return flow 13.4mm/d)
27
26
25
24
23
22
21
20
19
18
17
16
9-Jul 10-Jul 11-Jul 12-Jul 13-Jul 14-Jul 15-Jul 16-Jul 17-Jul 18-Jul
0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00
27
26
25
24
23
22
21
20
19
18
17
16
8-Aug 9-Aug 10-Aug 11-Aug 12-Aug 13-Aug 14-Aug 15-Aug 16-Aug 17-Aug
0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00
Fig. 12 Calculated water temperature variation at the plot G which is affected by both climatic factor and
return flow ((a): 9 – 17 July 2011, (b): 8 – 16 August 2011)
5. Conclusions
In this study, water temperature variation in dual-purpose canal was precisely observed along the
main canal of Tedori-gawa Shichika irrigation area in Japan. The water level and temperature variation
of the branch drainage channel was also investigated. As a result of field observation, it has become
clear that there is a large increase in water temperature relative to flow distance especially in
downstream part. The water temperature increase is thought to be due to both climate factors and
return flow from paddy fields.
Then a water temperature simulation model in dual-purpose canal was proposed which is based
on heat advection with water flow and heat balance in open channel. As a result of calculation by the
model, the amount of return flow from paddy fields was estimated. Then, the simulation model was
improved by incorporating the effect of return flow and the verification of the model was done by
comparing calculation results with observational data.
Acknowledgements
This study was supported by Ministry of Education, Culture, Sports, Science and
Technology-Japan through “Research Program on Climate Change Adaptation”. The authors would
like to acknowledge the field observation assistance of Tedori-gawa Shichika land improvement
district. The authors are also grateful to Dr. Tadashi Tsukaguchi in Ishikawa Prefectural University and
Prof. Masaru Mizoguchi in the University of Tokyo for providing the meteorological data. Special
thanks are due to the farmers who allowed us the observation in their paddy fields.
References
Iwakiri, S., 1964, Studies on the variation of heat balance characteristics of water layer under plant
cover in relation to the luxuriance of rice plant, Journal of Agricultural Meteorology, 19(3),
89-95 (in Japanese with English summary).
Kondo, J., 1995, Diurnal temperature variation of the river water (1) Model, Journal of Japan Society
of Hydrology & Water Resources, 8(2), 184-196 (in Japanese with English abstract).
Morita, S., 2008, Prospect for developing measures to prevent high-temperature damage to rice grain
ripening, Japanese Journal of Crop Science, 77(1), 1-12 (in Japanese with English abstract).
Nagahata, H., Nakamura, K., Ino, M., Kuroda, A. and Hashimoto, Y., 2005, The cultivation
management to make the occurrence of the milky white kernel and the cracked rice reduce under
high temperature during the ripening period, Bulletin of the Ishikawa Agriculture Research
Center, 26, 1-10 (in Japanese with English summary).
Nakagawa, H.,2013, High-temperature damage to grain ripening in rice, Water, Land and
Environmental Engineering, JSIDRE, 81(4), 52 (in Japanese).
Oue, A. and Kamii, Y., 2002, Research work on the water and heat balance of a paddy field, Research
Reports of Kochi University, Agricultural Science, 51, 45-76 (in Japanese with English abstract).
Tashiro, T. and Ebata, M., 1975, Studies on white-belly rice kernel IV. Opaque rice endosperm viewed
with a scanning electron microscope, Japanese Journal of Crop Science, 44(2), 205-214 (in
Japanese with English summary).
THE FUTURE
F-05
The Suitability Evaluation of Dredged Soil from Reservoir as

Embankment Materials
Jaesung Park, Younghwan Son, Sookack Noh, Taeho Bong

Rural Systems Engineering Department, College of Agri. and Life Sciences Seoul National
University
ABSTRACT
There are about 17,000 agricultural reservoirs, which have performed important role in supplying
agricultural water in South Korea. In order to stable water supply, it is necessary to maintain storage
capacity by dredging the sediment of reservoir. Most dredged sediment soil have been landfilled other
sites as wastes. In this study, to recycling the dredged soil, physical properties and engineering
characteristic as well as environmental stability were investigated. The 3 site samples were taken from
Mulwang(MW), Ansung(AS) and Jechon(JC) reservoir. To determine the environmental stability and
geo-chemical properties, the degree of heavy metal contamination, pH, EC, XRF and XRD analysis
were carried out. Also to evaluate recycling possibility as embankment materials, geotechnical
characteristics were investigated such as compaction characteristic, permeability test, direct shear
test and triaxial compression test. As a result, the contents of heavy metals were detected less than the
statutory standard. The test results of soil mechanics properties show that the dredged soil can be
used as fill-dam embankment materials.
Keywords: dredged soil, reservoir, embankment, sediment, material properties, heavy metals
1. Introduction
There are about 17,000 agricultural reservoirs in South Korea. The reservoir is recognized as a
core element of agricultural infrastructures due to the paddy farming is a largest part of Korean
agriculture. The primary purpose of the reservoir for agriculture farming is a stable supply of water.
However it was suffered drought from winter to early spring while floods occur in the summer
because of distinct seasonal difference of rainfall. In this reason, the project of embankment raising
had been conducted to stable supply of agricultural water and reducing flood damage caused by heavy
rains in summer. This project is expected to continue in the future, which is required to embank a large
amount of material.
On the other hand, reservoirs continuously have soil and sediment deposit coming from the
upstream, which causes reducing of water storage capacity of lake. Thus, the reservoir sediment
dredging process has been conducted on a regular. However, these reservoirs dredged soils have been
disposed of in the landfill yard as waste. This is the reason that many researchers have been studied to
find out the way to recycling the sediment dredged soils. Nevertheless, limitations of recycling are still
present due to lack of characteristic information.
The objectives of this study were as follows: (1) To determine the concentration of heavy metal
and geochemistry properties of dredged soil from reservoirs, (2) to evaluate the possibility of using for
embankment materials by various engineering properties test and (3) to comprehensive safety
assessment of fill-dam using dredged soils as embankment material by numerical simulation method.
2.1 Materials
The dredged soil materials used in this study are obtained from two reservoir managed by city
administration-Ansung(AS) and Jechon(JC) and one reservoir-Mulwang(MW) managed by Korea
Rural Community Corporation(KRC). The location of sampling sites is shown in Fig. 1.
Mulwang(MW)
Jechon(JC)
Ansung(AS)
Fig. 1 Locations of sampling sites
The dredged soil sample of Mulwang(MW) reservoir managed by KRC was collected in dry season by
shovel while Ansung(AS) and Jechon(JC) sediment soil were sampled in filed for discard. The surface
layer (0~50 cm) of the sediments was collected and at each site, 5~10 sub-samples were obtained and
mixed to yield a composite sample. The results of fundamental properties of three site samples are
shown in Table.1 and the particle size distribution curve is shown in Fig. 2
Table 1 The physical properties of samples
Sample Specific Gravity Cu Cg LL PL USCS

(Gs) (%) (%)
AS 2.60 5.88 1.30 43.7 37.6 ML
JC 2.56 44.89 0.23 47.6 37.0 SM
MW 2.56 14.00 1.41 33.6 25.6 ML
Cu : Coefficient of uniformity
Cc : Coefficient of curvature
Fig. 2 The curves of particle size distribution of samples
2.2 Geochemical properties and Heavy metal contents

Geochemical characteristics of the study samples were analyzed by pH, electrical conductivity
(EC), XRF and XRD. Soil pH(ISO -10390, 2005) and EC(ISO-11265, 2005) were determined by
adding water to dry soil (1:5 mixture) and measuring the pH and EC after 30 minutes by employing
pH meter and EC meter. The mineralogy of sediment samples was determined using X-ray diffraction
(XRD) and major element concentrations of dredged soil samples were determined using X-ray
fluorescence (XRF). The elemental concentrations are expressed as the percent of oxide mass,
corrected for the losses during ignition (heating the sample material at 1000 °C for 1 h in order to
eliminate volatile compounds) The microscopic structure of dredged soil was acquired by the SEM
technique Also to analyze the environmental safety of dredged soil, heavy metal contents were
measured by ICP-AES and ICP-MS
2.3 Engineering properties

To investigate the basic engineering properties of soil, standard compaction test and
permeability test were performed. It was calculated the maximum dry unit weight (γdmax), optimum
water content (OMC) and permeability coefficient. In order to estimate the shear strength of each
material, direct shear test and CU triaxial test were performed.
2.4 Suitability assessment

To evaluate the suitability of dredged soil for embankment materials, the strength properties and
coefficient of permeability of samples were used to numerical simulation. The slope/w and seep/w
module of GEO-STUDIO were used to assessment slope stability and seepage stability.
3. Results & Discussions
3.1 pH and EC
Table.2 displays the pH value and EC value of each site samples. The results showed that all
the samples exhibit slightly acidic (4.34~5.37). An EPA regulation (CFR, 2004) states that solid waste
exhibits the characteristics of corrosivity if a representative sample of the waste is aqueous and has a
pH less than or equal to 2(very acidic) or greater than or equal to 12.5(very alkaline) According to the
corrosivity criteria with respect to pH value, the sediments would be classified as non-corrosive. EC
value range is 93.9~253.3 μS/cm, which is similar to those of drinking water. According to the FAO
report (Ayers, 1989) and Bauder (2010), the recommend EC value for irrigation water is less than
0.7~0.75dS/m(dS/m=106 μS/cm). So, this soil is not expected to have a salt problem.
Table 2 The results of pH and EC of testing samples
Sample pH EC (μS/cm)
AS 4.34 164.1
JC 5.37 93.9
MW 5.22 253.3
3.2 Major chemical element and mineralogy

The result of XRF analysis was shown in Table.3. The results show that the major chemical
elements are SiO2, Al2O3 and Fe2O3, which is very similar to normal weathered and residual soil.
Especially, the loss on ignition (L.O.I) value range is 1.99~3.43%, slightly low. Also AS and JC soil
have more portion of Al2O3 than MW sample, while MW have more portion of Fe2O3.
Table 3 The each percentage of major chemical elements of samples from XRF analysis
Sample SiO2 Al2O3 Fe2O3 TiO2 MgO CaO Na2O K2O MnO P2O5 L.O.I
AS 53.48 25.91 9.41 1.16 2.12 0.93 0.49 3.81 0.11 0.14 1.99
JC 54.82 24.37 8.01 1.05 1.82 1.57 1.04 2.89 0.14 0.39 3.43
MW 52.96 21.69 11.64 1.33 1.58 2.22 0.63 4.13 0.21 0.55 2.22
The Fig.3 shows that the results of XRD analysis of soil samples. The main mineralogy is a
quartz for all samples. It is shows that the sediment samples are similar to normal residual soils, too
These results implies the sediment dredged soils are expected to have similar engineering
property to those of residual soils
Fig. 3 The results of XRD analysis
3.3 Environmental safety – Heavy metal contents
Mostly, the reservoir sediment transported by rainfall or stream from near the watershed. Thus
the sediment may have the hazardous substance from nearby factory or barn. For this reason, in this
study, heavy metal content of 8 kinds (Cd, Cu, As, Hg, Pb, Cr, Zn and Ni) which were controlled by
law was analyzed. The result of heavy metal contents is shown in Table.4 Generally Zn ion was most
detected while Cr6+ ion was not detected in all samples. Also, Ni, Pb and Cu ion were relatively more
detected than other heavy metal ions. MW samples have more Pb than any other samples. Although
the samples have some heavy metal ions, the value is under the “concern criteria” of the Korean Soil
Environmental conservation Act (for soil). It means that the dredged sediment can be used as a
recycling material.
Table 4 The heavy metal contents of soil samples
Sample Cd Cu As Hg Pb Cr6+ Zn Ni
AS 1.00 22.87 6.80 0.557 15.93 ND 81.60 34.50
JC 1.03 5.70 7.13 0.242 14.43 ND 76.30 7.80
MW 1.03 18.97 5.97 0.227 47.20 ND 93.30 19.07
Concern Criteria 4 150 25 4 200 5 300 100
Control Criteria 12 450 75 12 600 15 900 300
3.4 Compaction characteristic and Permeability

The results from standard compaction test and falling head permeability test are shown in
Table.5. Fig.4 showed compaction curves of test samples. Test results show that the maximum dry
density of soils are 16.0~16.3kN/m3, value deviation is not greater between the three testing soils.
Optimum moisture contents (OMC) of each samples is 17.6~21.9%. The permeability coefficient
values using the specimens have more than 90% relative compaction of γdmax are about 10-7 cm/s,
indicating a range similar to those of silty sand or silt.
Fig. 4 Compaction curves of soils
Table 5 The compaction property and permeability of samples
Sample γdmax (kN/m3) OMC(%) Permeability coefficient(k, cm/s)

AS 16.0 21.9 2.26 X 10-7
JC 16.0 17.6 -
MW 16.3 21.1 1.00 X 10-7
3.5 Direct shear test

The shear stress and normal stress values during the direct shear test were calculated by
dividing the horizontal load and the normal load, respectively, by the corrected area of the specimen’s
cross section. (Muhunthan et al., 2004; Bardet, 1997) For each of the samples, the maximum shear
stress and the corresponding normal stress were obtained at each normal load. The results for each of
the three samples were plotted on a normal stress versus shear stress diagram. Fig.5 shows the results
of the direct shear test. The strength parameters, internal friction angle (ϕ) and cohesion (C), are
displayed at table 6.
Fig. 5 Shear strength plot for sediment sample by direct shear test
Table 6 The strength property results from direct shear test
Sample Internal friction angle(ϕ, ̊ ) Cohesion(c, kPa)

AS 37.2 19.4
JC 36.9 25.4
MW 38.8 19.6
Average 37.6 21.5
The result shows that the three reservoir sediment soil have very similar strength parameter.
The value of friction angle is 36.9~38.8 ̊ and cohesion is 19.4~25.4kPa. There was no significant
difference for each samples. The value of average, 37.6 ̊ friction angle and 21.5kPa cohesion, indicate
the deposit soil sample have the strength properties of sand or silt soil.
3.6 Triaxial compression test
The triaxial testing was accomplished in consolidated-undrained (CU) conditions. Test results
from each of these tests are presented Fig.6~8. All samples used in test were 50mm in diameter and
100mm in height. After the test samples were consolidated, the shear test was performed at confining
stress 50kPa, 100kPa and 150kPa with undrained situation. From the graph of shear stress-normal
stress, the failure envelope was drawn between 50kPa and 150kPa Mohr circles. That result shown in
Table. 7.
Table 7 The strength property results from triaxial test
Sample Internal friction angle( ̊ ) Cohesion(kPa)

AS 25.9 27.5
JC 27.1 3.5
MW 27.5 19.3
Average 26.8 16.8
Fig. 6 Shear strength plot for AS sample by triaxial compression test(CU)
Fig. 7 Shear strength plot for JC sample by triaxial compression test(CU)
Fig. 8 Shear strength plot for MW sample by triaxial compression test(CU)
4. Application to embankment material
4.1 Study reservoir

It was performed the applicability as embankment material of dredged soil using the strength
parameter from § 3.4~3.6. For this, the objective reservoir was selected. The reservoir selected in this
study, completed in 1949, is located on Han-river basin. This reservoir has a total storage capacity of
approximately 4.47x106 m3, a 2,873ha watershed size and a 927ha irrigation field. Embankment type
is a fill dam with zone core and the size is 17.9m height and 280m width. The cross section are shown
in Figure 9
Fig. 9 The cross section of the objective reservoir
4.2 Case study

In this study, following cases were used and the numerical analysis was carried out
Case 1 : Material substitution – embankment material

Case 2 : Material substitution – zone core
Case 3 : Raising the embankment using new material
(a) Case 1
(b) Case 2
(c) Case 3
Fig. 10 The concept drawing of 3 cases
4.3 Stability standard
The numerical analysis was performed according to the standard of design for fill dam of
Ministry of Agriculture (KMA, 2002). Table.3 shows the standard of safety factor for various cases. In
this study, usual conditions are concerned.
Table 8 The standard of the safety factor for fill-dam
Condition Safety factor This

Case Water level Seismic Study
of fill dam upstream downstream
1 Rapid down None 1.2 1.2 O
2 As usual Filled with water Concern 1.2 1.2 O
3 Usual level Concern 1.15 - O
4 Shortly after Bottom None 1.3 1.3 -
5 construction Usual level None 1.3 - -
4.4 Input data

For numerical analysis, the input data were chosen from original material properties of study
reservoir (KRC, 2008) and the strength parameter from § 3.4~3.6. The strength parameter of dredged
soil from triaxial test due to the parameters can be described well of situation of the study reservoir.
Table 9 The input data of numerical analysis for stability assessment
Material γt (kN/m3) ϕ(̊) C (kPa) k(cm/s) Source

-5
Embankment 18.48 25 17 7.83 X 10
Core 19.79 14 39 2.61 X 10-7 EAP plan of reservoirs.
Soft rock 24.53 35 100 1.00 X 10-9 (KRC, 2008)
Surface rock 18.15 40 0 0.1
Dredged soil 18.55 26.8 16.8 1.00 X 10-7 In this study
4.5 Results of simulation

The results of stability simulation are shown in Table 10. All cases and water level situation
are satisfied the standard of safety factor. Fig.11 shows the example of analysis results for Case 3
Table 10 The results of numerical analysis

Situation Case 1 Case 2 Case 3 Standard
1. Rapid down 1.2
Upstream 1.612 1.882 1.659
Downstream 2.077 2.027 1.839
2. Filled with water 1.2
Upstream 2.768 2.554 2.413
Downstream 2.077 2.027 1.709
Upstream(/w seismic) 1.528 1.404 1.338
Downstream(/w seismic) 1.441 1.415 1.304
3. Usual level 1.15
Upstream 2.141 2.030 1.862
Upstream(/w seismic) 1.395 1.327 1.227
1.227
1.659
60
60
50
50
40
Elevation
40
Elevation
30
30
20
20
10
10
0
0 20 40 60 80 100 120 140 160 180
0
0 20 40 60 80 100 120 140 160 180
Distance
Distance
(a) Rapid down water level at upstream (b) As usual water level at upstream(/w seismic)
1.304
1.709
60
60
50
50
40
40
Elevation
Elevation
30
30
20
20
10
10
0
0 20 40 60 80 100 120 140 160 180 0
0 20 40 60 80 100 120 140 160 180
Distance Distance
(c) Filled with water level at downstream (d) Filled with water level at downstream
(/w seismic)
Fig. 11 The numerical analysis examples for CASE 3
5. Conclusions
In order to evaluate suitability of dredged sediment soils as embankment materials,

geochemical property, heavy metal contents and engineering property tests were performed for 3
reservoirs sample in this study. Also the suitability assessment of dredged soils as embankment
materials was carried out, too. The main conclusions are as follows.
1. The heavy metal contents of all dredged soil in this study are lower than the standard value,
which can be used as a recycling materials without a heavy metal contaminant problem.
2. The pH is slightly acidic, which is 4.34~5.37 and EC value range is 93.9~253.3 μS/cm,
which is can be used sufficiently as irrigation water without damage from salt water.
3. The major chemical element and mineralogy characteristic from XRF, XRD analysis show
that the dredged soil is very similar to normal weathered soil.
4. The strength properties of dredged sediment soil from reservoir were determined by
standard compaction test, falling heave permeability test, direct shear test and triaxial test. The soils
have γdmax of 16.0~16.3kN/m3, a permeability coefficient of about 1.0x10-7cm/s, average friction
angle of 26.8 ̊ and cohesion of 16.8kPa.
5. To assessment of suitability for embankment materials, the numerical simulation for 3 cases
were performed. As a result, the dredged soil could be used not only substitution of embankment and
core material but also new embankment material for expansion of reservoirs with enough stability.
Acknowledgment
This paper was supported by the National Research Foundation of Korea (NRF) grant funded
by the Korea government (MEST) (No. 2012R1A1A1010633)
References
Ayers, R. S., and Westcut, D. W, 1989, FAO irrigation and drainage paper, Water quality for
irrigational purposes.
Bardet, J. P, 1997, Experimental soil mechanics, Prentice-Hall InternationalLtd.: Englewood Cliffs,
NJBauder, T. A., Waskom, R. M., and Davis, J. G. (2010): Irrigation water quality criteria.
CFR. Code of Federal Regulations, 2004, Title 40: Protection of Environment, Part 261: Identification
and Listing of Hazardous Waste Support C: Characteristics of Hazardous Waste, 261.22:
Characteristics of Corrosivity
ISO 10390, 2005, Soil quality - pH Difference in method
ISO 11265, 2005, Soil quality - Determination of the specific electrical conductivity
Korea Ministry of Agriculture. (KMA), 2002, The standard of fill-dam design.
Korea Rural Community Co.(KRC), 2008, The Emergencty Act Plan for Reservoirs
Muhunthan, B., R. Taha, and J. Said., 2004, "Geotechnical engineering properties of incinerator ash
mixes." Journal of the Air & Waste Management Association, 54(8),985-991.
THE FUTURE
F-08
Analysis of irrigation service needs by rice farming families

in Japan
Toshiaki Iida*, Masaomi Kimura*, Koshi Yoshida**, Naritaka Kubo*,

Takahiro Yokoi*
*Graduate School of Agricultural and Life Sciences, The University of Tokyo, Japan
**College of Agriculture, Ibaraki University, Japan
ABSTRACT
As irrigation systems have been designed and managed as supply-oriented systems, the freedom of
water use by farmers has been restricted. On the other hand, the recent rapid development of ICT
enabled us to instantly transmit the information on the amount of available water and on the demand
of farmers interactively. Taking into these circumstances, it is considered that irrigation systems
should be managed as service providing systems for farmers and the service quality should be further
improved taking advantage of recent ICT. In the present study, the irrigation activities by farmers on
rice paddy fields were precisely observed to seek effective irrigation service for the farmers. Field
plots were selected near Aichi Canal in central Japan to precisely observe the water balance at the
plot and irrigation activities by the cultivating farmer. It was observed that a lot of water was drained
especially in the late cultivation period, probably caused by unnecessary irrigation attributed to
saving of labor for water management. While the farmers seemed to be interested in labor saving,
some of them were not willing to receive any new service to help their water management. It is
indicated that classification of farmers is necessary to consider effective irrigation service according
to the conditions of each farmer.
1. Introduction
From ancient times, irrigation systems have been generally designed and managed as
supply-oriented systems, as most of the irrigation projects have been carried out by a ruler in authority
who wants to strengthen the region by stable agricultural production making use of existing water
resources. Consequently, the degree of freedom for water use by individual farmers has been strongly
restricted by the amount of available water at the headwork and conditions of flowing water through
the canal network. In case of drought, farmers have often been obliged to wait for water or even to
give up their crops without any information on expected water supply.
On the other hand, owing to recent rapid
development of information communication 130o E
technology (ICT), it has become possible to instantly
transmit information on the amount of available water
and on the demand of farmers interactively, making The Japan Sea
use of Internet and various field sensors (Olalla et al., 40o N
2003). Taking into these circumstances, it is
considered that irrigation systems should be managed Study area
as service providing systems for farmers and the
service quality should be further improved from a
demand-oriented point of view (Renault and
Montginoul, 2003).
Though a solid grasp of the demand is necessary
to improve service quality, the actual demands of 140o E
farmers in regard to their irrigation management have
not been scientifically investigated so far. It is Fig. 1 Location of the study area
necessary to grasp the real time activities for water
management which is considered to represent the
farmers’ demands. Only few attempts have so far been
carried out to precicely observe farmers activities in
their field from the viewpoint of service demand in
irrigation. In this study, irrigation activities by farmers P
on rice paddy fields were precisely observed in a
selected subject area in Japan to consider how to
provide the farmers with better irrigation service based
M
on a service dominant logic. Plot 1
M : Flow meter
2. Methods M a=command area
Plot 2
2.1 Site description Fig. 2 Experimental plots
One of the most downstream branches in Aichi
Canal irrigation system, located in central Japan, was
designated as the experimental area for the present study (Fig. 1). Aichi Canal project was completed
in 1961 to supply irrigation water to the agricultural area mainly in Chita peninsula. As the water
resources carried by Aichi Canal induced rapid industrial development as well as population increase
of the region, Aichi Canal second stage project was conducted during 1982-2004 to meet the changing
demands not only for agriculture but also for industry and domestic use in the region. In Aichi Canal
second stage project, go-ahead demand-oriented supply systems were partly adopted in Aichi Canal
irrigation system. Most of the secondary canals were converted from open channels to buried pipes
and novel upstream and downstream water levels’ control check gates were installed in the main canal
to increase water storage capacity of the main canal.
The experimental area is located in Handa city in Aichi prefecture (34o 54.2’ N, 136o 52.3’ E),
which consists of rice paddy fields. The irrigation water is pumped from the main canal of Aichi Canal
at Handa pump station to Handa branch pipe before gravitationally diverted to the tertiary pipe (Fig. 2).
As Handa pump station is operated from 8 to 17 o’clock, irrigation water is not supplied to Handa
branch during night hours, except special cases of severe drought. Among the rice paddy field plots
irrigated by the tertiary pipe, two experimental plots were selected for the present study as shown in
Fig. 2. The selected experimental plots are called Plot 1 and Plot 2 hereafter. The inlet valve of
irrigation water to the experimental plot was manually operated by the farmer who cultivates the plot.
The characteristics of the farmers cultivating the selected eperimental plots are given in Table 1.
2.2 Measurements
The discharge of irrigation water to the experimental plot was measured at every 10 minutes by
an electromagnetic water meter (Aichi tokei denki Co., Ltd., SA075GS) at the water inlet of the plot.
The photograph around the water outlet of the plot was taken at every 1 hour automatically by a field
camera (Brinno Inc., Garden Watch Cam) to observe the depth of the floodwater, the floodwater color,
and the conditions of rice plants on the experimental plots. On the occasion of every water
management action, the time and the opening level of the inlet valve were recorded by the farmer.
Other farming activities were also recorded by the farmer throughout the rice cultivation period. The
cultivation calendar of rice plants in 2012 at each experimental plot was shown in Table 2. The
meteorological data observed by Japan Meteorological Agency at nearby stations and the precipitation
data observed by Aichi Canal land
improvement district were referred to.
Table 1 Characteristics of the farmers cultivating
2.3 Water balance at a plot the selected plots
The daily water balance at each Plot 1 Plot 2
experimental plot was calculated for the whole Involvement in farming Full-time Part-time
cultivation period. Because the farmer Farming experience (y) >40 >40
cultivating Plot 1 mentioned that a lot of lateral Number of workers 2 2
seepage between plots occurred, the water
(persons)
balance at a plot was evaluated taking lateral
seepage into account, using the equation Distance from the 2.1 2.4
residence (km)
Other information Retired person
I + Pi + R = ET + D + Po + dH farming for (1)
enjoyment
where I is the amount of irrigation water

(mm/d), Pi is the lateral seepage inflow from the
adjacent plot (mm/d), R is the precipitation Table 2 Cultivation calendar
(mm/d), ET is the evapotranspiration (mm/d), D is Plot 1 Plot 2
the amount of surface drainage water (mm/d), Po Puddling May 23 Apr. 30
is the sum of the lateral seepage outflow to the Transplanting May 29 May 5
adjacent plot and the vertical percolation (mm/d),
Mid-summer from Jun. 29 Jun. 14
dH is the changes in the floodwater depth (mm/d).
Eq. 1 is rearranged in drainage to Jul.16 Jul. 14
Full drainage Sept. 30 Aug. 19
Harvest Oct. 7 Sept. 16
(Po + D) – Pi = I + R – ET – dH (2)
for evaluation of the inflow and outflow of a plot. As the soil surface was continuously flooded or wet
enough, the actual evapotranspiration was considered to be equal to the potential evapotranspiration
calculated by Penman method. The albedo used in Penman method was obtained based on Kotoda
(1986). dH was obtained by analyzing the photograph image around the water outlet to detect the
changes in the floodwater depth.
2.4 Interview with the farmers

Interviews were conducted with the farmer cultivating Plot 1 and Plot 2. A farmer and 3
interviewers seated around a table and talked each other for about 2 hours. In the interview, details of
water management activities at their own paddy fields as well as whole agricultural management in
their farm household were asked. The possibility for them to recieve new service supporting their
irrigation management was also investigated.
mid-summer
drainage Precipitation
P T Irrigation F
120
Depth (mm/d) Percolation (Into)
80 (a)
40
0
5/1 6/1 7/1 8/1 9/1 10/1
0
Depth (mm/d)
40 (b)
80
Evapotranspiration
120 Percolation (Out) + Drainage
5/1 6/1 7/1 8/1 9/1 10/1
Opening level
full
middle (c)
little
shut
5/1 6/1 7/1 8/1 9/1 10/1
Date (2012)
Fig. 3 Water balance ((a): Inflow, (b): Outflow) and the opening level of the inlet valve (c) at Plot 1,
April 27 – September 30, 2012. P : Puddling, T : Transplanting, F : Full drainage.
3. Results
3.1 Observation in Plot 1

Fig. 3 (a) and (b) show the obtained water balance at Plot 1 for the cultivation period in 2012.
Before the mid-summer drainage, irrigation water was supplied only at the time of puddling and
transplanting. In spite of no irrigation water supply from the transplanting on May 29 to the beginning
of the mid-summer drainage on June 29, almost stable floodwater depth was kept. It was revealed in
an interview with the farmer that the floodwater in Plot 1 was sustained by lateral seepage from the
adjacent plot which was about 0.5 m higher than Plot 1. At the beginning of the mid-summer drainage
on June 29, the floodwater of 8 cm in depth was completely drained. During the mid-summer drainage,
no irrigation water was supplied while drainage occurred responding to precipitation. After the
mid-summer drainage, drainage occurred in case water was supplied more than 50 mm/d by
precipitation or irrigation. Fig. 3 (c) shows the opening level of the inlet valve at Plot 1 along with
time. Although the inlet valve was operated at the time of puddling and transplanting, it had not been
operated until the end of the mid-summer drainage owing to lateral seepage from the adjacent plot, as
mentioned above. After the mid-summer drainage, the inlet valve was intermittently opened for 2 – 6
hours with intervals of 2 - 14 days.
mid-summer
PT drainage Precipitation
Irrigation
Depth (mm/d) 200 Percolation (Into)
(a)
F
100
0
5/1 6/1 7/1 8/1 9/1 10/1
0
Depth (mm/d)
100
(b)
Evapotranspiration
200 Percolation (Out) + Drainage
5/1 6/1 7/1 8/1 9/1 10/1

Opening level
full
middle (c)
little
shut
5/1 6/1 7/1 8/1 9/1 10/1
Date (2012)
Fig. 4 Water balance ((a): Inflow, (b): Outflow) and the opening level of the inlet valve (c) at Plot 2,
April 27 – September 30, 2012. P : Puddling, T : Transplanting, F : Full drainage.
3.2 Observation in Plot 2
Fig. 4 (a) and (b) show the obtained water balance at Plot 2 for the cultivation period in 2012.
Different from the case of Plot 1, irrigation water was intermittently supplied after the puddling on
April 30. Until the beginning of the mid-summer drainage, drainage water was minimal and stable
floodwater depth was kept by intermittent supply of irrigation water. During the mid-summer drainage,
no irrigation water was supplied, while drainage occurred responding to precipitation. On the other
hand, a lot of water was drained after the mid-summer drainage, principally caused by unnecessary
irrigation after the mid-summer drainage. Fig. 4 (c) shows the opening level of the inlet valve at Plot 2
along with time. Before the mid-summer drainage, the opening duration of the inlet valve was shorter
than that after the mid-summer drainage, indicating more frequent water management activity by the
farmer in the early stage of rice cultivation. After the mid-summer drainage, the inlet valve kept
opened and closed for several days repeatedly. It is considered that the considerable volume of
drainage water after the mid-summer drainage was attributed to the less frequent handling of the inlet
valve which brought about unnecessary irrigation water.
3.3 Interview results
Part of the interview results regarding the possibility to recieve new service is shown in Table 3.
Both of the farmers cultivating the experimental plots were reluctant to accept new service, because
they are afraid of possible yield decrease by entrusting the cultivation work to other people. It seemed
very hard for them to accept any additional cost for new service. On the other hand, both farmers
pointed out that agricultural corporations or farmers with large-scale fields may have needs for various
services to help the farming work.
4. Discussion
4.1 Actual conditions of the farmers

Because there was a normal precipitation amount during the cultivation period in 2012, it is
considered that the observed data represent the situation in normal years. As different conditions must
be expected in drought years, the following discussion can be applied to normal years only.
Because Handa pump station is run from 8 to 17 o’clock and is stopped during night hours,
irrigation water does not come during night even if the inlet valve at each paddy field plot is opened. It
is interesting that the inlet valve was kept opened for a couple of days at Plot 2 despite that irrigation
water does not come especially after the mid-summer drainage, probably in order to save labor for
irrigation management.
Table 3 Results from interviews with the cultivating farmers

Answers
Questions
The farmer cultivating Plot 1 The farmer cultivating Plot 2
Opinion about outsourcing of The farming is one of my pleasures. I I am afraid of possible yield
water management will go to the fields every day, even if decrease by outsourcing. I want to
farming work is outsourced. I think grow my rice by my way.
better yield can be obtained if I grow rice
by myself.
Concrete work which can be Fertilizing Check of water leakage
possibly outsourced
Acceptable cost for new No additional cost is acceptable No additional cost is acceptable
service
At the same time, it is considered that less frequent handling of the inlet valve after the
mid-summer drainage has lead to much drainage water in Plot 2. Because the flashboard was fixed at
the outlet of the plot, much drainage water means unnecessary irrigation water unless there were other
purposes such as water temperature control. It is probable that the farmers are not so much interested
in saving the amount of water because the irrigation water fee is charged by the area of the plot, not by
the volume of consumed water, in Japan.
However, unnecessary irrigation water may lead to increase in the electricity fee for Handa
pump station. Because all the tertiary canals diverted from Handa branch are influenced by the
operation of Handa pump station, it is needed to evaluate the effect of runtime cut of Handa pump
station on a share of electricity expense as well as irrigation water supply conditions to each tertiary
canal.
It is considered that the intermittent run of Handa pump station and the water fee charged by the
area facilitated saving of labor for water inlet handling by farmers although a considerable amount of
irrigation water was ineffectively drained. It is suggested that effective irrigation service can be
proposed on the basis of more precise study on the relation among the water fee, the labor for the inlet
valve handling, and the share of electricity expense for Handa pump station in this tertiary canal.
4.2 Possible concrete service provision for farmers
The interview results showed that the farmers cultivating the experimental plots were negative
to receive new service, because they are afraid of possible yield decrease by entrusting the cultivation
work to other people and are not willing to pay additional cost to change their accustomed work. The
reluctance may come from their unwillingness to change their farming activities as they are satisfied
with the current stable farming style.
On the other hand, the farmers pointed out that agricultural corporations or farmers with
large-scale fields may have needs for various services. It is indicated that effective service varies
widely depending on the conditions of the farmer being offered the service. It is strongly suggested to
be necessary to classify farmers with respect to the service needs in order to consider appropriate
irrigation service to each farmer.
The precise observation of irrigation activities in the present study revealed that the farmers are
much interested in saving labour for irrigation management. The tendency was also expressed in the
interview. Moreover, it was grasped in the interview that the main observation point by the farmers in
their daily patrol of their plots is the depth of flood water and the condition of crops. As the
information of the flood water in each plot is crucial for farmers, it is strongly suggested that remote
provision of the flood water condition through information network can be possible concrete service
for farmers. Owing to such a service tool, farmers can decide whether they go to their plot for
irrigation management by their portable terminal anywhere. The needs for the flood water condition
provision service must increase in future, because if the total area cultivated by one farmer expands in
future, the labour for the daily patrol also increases. In order to evaluate the feasibility of such service,
the cost and the benefit of the service provision should be carefully estimated, taking the water fee and
the electricity expense for pump stations into account.
5. Conclusions
The water balance of a paddy field plot and water management activities by farmers were
investigated at the experimental plots in Aichi Canal irrigation project located in central Japan to seek
effective irrigation service for farmers. A lot of water was drained after the mid-summer drainage,
probably caused by unnecessary irrigation attributed to saving of labor for water management. While
the farmers seemed to be interested in labor saving, they were not willing to receive any new service
to help their water management. It is suggested that more precise evaluation of the relation among the
water fee, the labor for the inlet valve handling, and the share of electricity expense for the pump
station may provide useful information. It was also indicated that classification of farmers is necessary
to consider effective irrigation service according to the conditions of each farmer.
As the flood water is the main observation point by farmers in their daily patrol, remote
provision of the flood water condition through information network can be valuable service for
farmers. The feasibility of such service should be evaluated, taking the water fee and the electricity
expense for pump stations also into account.
Acknowledgments
This study was supported by Research Institute of Science and Technology for Society in Japan
Science and Technology Agency through “Service Science, Solutions and Foundation Integrated
Research Program”. The authors are grateful to staff members of Aichi Canal land improvement
district for their assistance in the field observation. Special thanks are due to the farmers who allowed
to carry out this study in their rice paddy fields.
References
Kotoda, K. (1986), “Estimation of river basin evapotranspiration.” Environmental Research Center

Papers 8, University of Tsukuba, 66.
Olalla, F. M. de S., Calera, A., Domı́nguez, A. (2003), “Monitoring irrigation water use by combining
Irrigation Advisory Service, and remotely sensed data with a geographic information system.”
Agricultural Water Management, 61, 111-124.
Renault, D., Montginoul, M. (2003), “Positive externalities and water service management in
rice-based irrigation systems of the humid tropics.” Agricultural water management, 59, 171-189.
THE FUTURE
F-10
Runoff Characteristics of Non-point Source Pollution from

Reclaimed Paddy
Yujin Lee, Chun Gyeong Yoon, Joon-Sik Kim, Moonsoo Cho, Seungil Lee
Department of Environmental Science, Konkuk University, Seoul, Korea
ABSTRACT
In Korea, since the 1960s, a reclamation project to steadily reclaim areas has made significant
progress. A reclamation project that was conducted earlier in an area with a lot of food production for
the purpose of financing has been used as agricultural land. Farming techniques of various forms
exist and tillage methods depend on soil and local, climate characteristics; and the type of
contaminants, irrigation and fertilization. Accordingly, because the agricultural non-point pollutant
source overflow rate largely depends upon the rainfall amount as well as various agricultural
activities, an investigation of rice paddy overflow is required, considering weather, farming, water
quality management, and environmental change. Even though there would be difficulty in quantization
that is reproducible because of the complexity of the characteristics of agricultural non-point
pollutant sources, a close investigation and monitoring of the loading characteristics of pollutant
materials overflowing from the rice paddy are needed. This study was performed to investigate the
characteristics of agricultural non-point source pollutant irrigation at paddy fields of Seokmoon,
Dangjin. To investigate nutrient balance, the monitoring was accomplished over 5 months from June
2011 to September 2011. The specimens taken from the subject areas were analyzed for 10 water
quality parameters in total: BOD, COD, TOC, SS, TN, NH3-N, NO2-N, NO3-N, TP and PO4-P,
according to the Standard Method (APHA). For the water balance, the monitoring of the study area
was conducted for five months (6~10, 2011) during the paddy rice farming season in Korea. During
the investigation period, rainfall was 1185.5 mm, infiltration 30.4 mm, evapotranspiration 260.1 mm,
and the irrigation water of the block scale 4071.3 mm and paddy scale 544.9 mm. The pollution load
in the block scale was 6.8, 26.1, 5.9, 44.3, -0.3 and -1156.6 kg/ha and in the paddy scale -5.9, 43.2,
6.8, 2.2, 3 and 213.6 kg/ha for BOD, COD, TOC, TN, TP and SS, respectively.
Keywords: Non-point source, water balance, mass balance, reclaimed paddy
1. Introduction
Water quality policy in Korea has usually focused on point sources, which includes sewage and
factory wastewater. This has created a fundamental limitation in that the regulations were not
sufficient enough to improve the water quality of rivers and reservoirs. The current sewer rate is
91.8%, based on 2012 readings. And mostly the sewage treatment plant has been complete the
advanced processing system. However, it is not at a level that everyone is satisfied with. Because
non-point source pollutants have been rapidly increasing, many scientists have carried out research on
them. However most of the research has investigated urban areas. In Korea, agriculture land occupies
25% of the total land area, which is larger than the urban land area (6.2 % of total land). But most of
the focus on the urban land is non-point source. Paddy fields occupy over 60 % of total farming lands.
The nutrient balance of the paddy fields controls the nutrient supply and demand for rice plants and
has an impact on the water environment. The environmental capacity of agricultural non-point sources
among several non-point sources such as pesticides, fertilizers and agricultural drainage—which can
have an impact on the water environment—has not been evaluated properly. Thus, non-point source
pollution loads of nutrients are suspected to be a major contributor to the problem and must be
addressed. Hydrologic phenomenon and runoff characteristics of the agriculture land are substantially
different from other land uses. Consequently, applying the existing research results is difficult.
Therefore, quantitative evaluation is required.
The nutrient balance and water balance of agricultural land differ by farming method, for
example the physical features of soil, water source, water management methods, and so on. This paper
investigates the circulation mechanism of non-point source pollutants from paddies, which account for
more than 60% of agricultural water sources. A background data survey was conducted on the
historical weather data; water quality of water sources; and physical characteristics of the monitoring
point of water quality, soil and climate; as well as fertilization, irrigation, and drainage management.
Then a field monitoring study was carried out to investigate the water balance and nutrient balance of
paddy fields. The experimental field was investigated on 1 June, 2011 to 30 October, 2011. And in
order to consider the characteristics of rice paddy pollution overflow, rainfall analysis was conducted
using rice paddy data. Then the data were analyzed by comparing the concentration of pollutant, water
balance and nutrient balance. Since monitoring was performed for only five months, it was too
difficult to determine the general characteristics of non-point sources in the reclaimed land.
Therefore, it was important to establish a non-point source database, which has a high
confidence level, and various research related parameters through continuous monitoring.
2.1 Study site

The study paddy field site was
located in Seokmoon polder
(Seulhang-ri, Godea-myun, Dangjin-Si,
Chung cheongnam–do) watershed in
Korea (Fig. 1). This rice cultivation area
had been protected from livestock
farming, a farmstead and an industrial
complex. The area has a moderate
climate, with an average annual
temperature of 12.0 ℃, ranging from
33.5 ℃ in August to -13.1 ℃ in Fig. 1 Location of research facilities
January. Average annual rainfall is
909.6 mm, over 50% of which falls in July and August. The field work was carried out during an
irrigation season from June to October 2011 in Korea. The paddy field site has an area of 216,000 ㎡
and is divided into 12 paddy fields. The field was supplied with water through a pipe connected to the
drainage channel and has an output pipe of 1~2.
2.2 Farming Practices
The experimental field was first plowed on 8 June, and irrigated from 9 June. The field had a
second plowing and fertilization from 9 June to 11 June. The fertilizer was dispersed in all layers of
the chemical fertilizer 134.0 kg N /ha, 102.1 kg P/ha and 63.2 kg K/ha. And the rice transplant
distance was 15 cm x 30 cm four seeding from 20 June to 22 June and the rice was harvested from 24
October to 31 October.
2.3 Water balance analysis

Water flow was continuously measured in each paddy field during the study period. The spot
measurement of the inlet and outlet flows of each paddy field was made using small current meters.
Flows were also monitored and recorded using pressure-type water gauges and a data logger to assist
in spot measurement and verify continuous flow. The water balance was determined to assess the
treatment efficiency of the constructed paddy field. The water balance of the field was calculated with
the following equation:
Wi = Wi − 1 + Ini + Raini − (Outi + ETi + Infi) (1)
Where, Wi: water balance in I day (mm), Wi-1: water balance in i-1 day (mm), Ini: influent
(mm), Raini: precipitation (mm), Outi: outflow (mm), ETi: evapotranspiration (mm), Infi: infiltration
(mm)
Infiltration and evapotranspiration were measured and calculated with meteorological data.
Evapotranspiration was determined with the FAO Penman-Monteith equation.
2.4 Nutrient balance analysis

was twice a month during the study period of June 2011 to October 2011. Conventional water
quality parameters including DO, TOC, BOD5, CODCr, Total-N, NO3-N, NO2-N, NH4-N, Total-P,
PO4-P, pH, and total suspended solids (TSS) were analyzed using standard methods (APHA, 2005).
The constructed paddy performance was determined by the following equation:
Ii + IPR = ODR + OINF + OHRV (2)
OnET = ODR + OINP ( II + IPR) (3)
Where, Ii: inflow of water, IPR: rainfall inflow, IPER: inflow fertilizer, ODR: drainage runoff,
OINF: infiltrate runoff, OHRV: runoff through the harvest of plant, ONet: net pollutant loads
Furthermore, if ONet was (-) that the paddy acted as a purification system and if ONet was (+) that
the paddy acted as a discharge system.
3. Result and Discussion
3-1. Water balance

In the investigation period, total precipitation was 1185.5 mm, irrigation 4071.3 mm, runoff
4833.7 mm, infiltration 30.4 mm, and evapotranspiration 260.1 mm. Total influent (sum of irrigation
and precipitation) was 5256.8 mm and water loss (sum of runoff, infiltration and evapotranspiration)
was 678.5 mm. The highest irrigation was observed in mid-June at 724.0 mm, and lowest irrigation
was observed in early August. Irrigation may be reduced due to midsummer drainage. Outflow
increases or decreases in proportion to the amount of rainfall and irrigation. The amount of infiltration
was very small and can be ignored. Outflow was increased in the winter season due to ice formation.
The low effluent load could be underestimated because of the low flow rate as well as the low nutrient
removal rate in winter.
Fig. 2 Water balance of paddy field
Table 1 Water balance
Inflow (mm) Outflow (mm)

Time
Rainfall Irrigation Subtatoal Runoff Infiltration Evapotranspiration Subtatoal
6/M 0.0 724.0 724.0 640.4 3.2 34.9 678.5
6/L 400.9 465.0 865.9 697.1 4.2 21.7 723.1
7/F 169.0 269.7 438.7 436.9 3.1 23.4 463.4
7/M 225.2 580.3 805.5 799.3 3.1 30.5 832.9
7/L 180.4 335.8 516.2 564.4 3.1 24.0 591.5
8/F 62.5 31.2 93.7 54.6 1.5 10.4 66.5
8/M 99.5 44.2 143.7 130.5 2.5 14.6 147.5
8/L 1.0 255.0 256.0 237.4 2.5 25.9 265.8
9/F 1.5 448.6 450.1 401.2 2.3 25.2 428.7
9/M 18.5 458.5 477.0 434.2 2.4 24.1 460.7
9/L 27.0 458.5 485.5 434.2 2.4 24.1 460.7
10/F 0.0 0.5 0.5 3.4 0.2 1.3 5.0
Total 1185.5 4071.3 5256.8 4833.7 30.4 260.1 678.5
3-2. Nutrient balance
Fig. 3 Temporal variation of BOD5 and CODCr concentration
The average BOD concentration of the inflow was 4.08 (1.2 ~ 9.6) mg/L and the outflow was
4.03 (0.9 ~ 13.2) mg/L. The COD concentration of the inflow was an average of 25.2 (14.4 ~ 38.5)
mg/L and the outflow was an average of 23.2 (8.9~48.0) mg/L. The average TOC concentration of the
inflow was 5.8 (3.1 ~ 8.4) mg/L and the outflow was 5.4 (1.9~11.5) mg/L. BOD, CODCr and TOC
were the largest concentrations in June because of the fertilization. The SS concentration of the inflow
was an average of 89.8 (18.0~137.2) mg/L and the outflow was an average of 56.7 (9.1~133.8) mg/L.
SS was increased after September due to the opening of the bank of the rice field and the increase in
soil.
Fig. 4 Temporal variation of TOC and SS concentration
Fig. 5 Temporal variation of nitrogen concentration
The average TN concentration of the inflow was 3.2 (2.2 ~ 4.8) mg/L and the outflow was 3.1
(1.9 ~ 7.8) mg/L. The NO3-N concentration of the inflow was an average of 1.7 (1.1 ~ 3.5) mg/L and
the outflow was an average of 1.7 (0.7 ~ 5.7) mg/L. The average NH4-N concentration of the inflow
was 0.38 (0.17 ~ 1.28) mg/L and the outflow was 0.37 (0.08 ~ 1.43) mg/L. The average NO2-N
concentration of the inflow was 0.61 (0.05 ~ 0.49) mg/L and the outflow was 0.16 (0.02 ~ 0.66) mg/L.
On the whole, nutrient N was highest in June and rose slightly in late August. That was fertilization,
too.
Fig. 6 Temporal variation of phosphorus concentration
The average TP concentration of the inflow was 0.36 (0.22 ~ 0.63) mg/L and outflow 0.34 (0.01
~ 1.19) mg/L. The PO4-P concentration of the inflow was an average of 0.17 (0.07 ~ 0.33) mg/L and
the outflow was an average of 0.14 (0.02 ~ 0.40) mg/L. The influent was the largest in June due to the
fertilization. And the plants were growing and absorbed a lot of phosphorus in July.
Table 2 Nutrient balance (TN)

Input (kg/ha) Output (kg/ha)
Time
Rainfall Irrigation Fertilizer Subtotal Runoff Infiltration Plant Subtotal
6/M 0.00 30.31 93.00 123.31 30.38 0.18 30.56
6/L 1.79 18.17 18.17 58.60 0.11 58.71
7/F 0.92 9.07 9.07 15.87 0.07 15.94
7/M 0.87 18.67 18.67 23.56 0.07 23.63
7/L 0.72 10.74 10.74 12.93 0.06 12.99
8/F 0.52 1.01 1.01 1.38 0.02 1.4
8/M 0.61 1.45 1.45 3.49 0.03 3.52
8/L 0.02 6.71 6.71 4.98 0.09 5.07
9/F 0.03 14.79 14.79 13.03 0.07 13.1
9/M 0.09 6.69 6.69 4.94 0.05 4.99
9/L 0.12 11.07 11.07 8.71 0.05 8.76
10/F 0.00 0.01 0.01 0.03 0.00 107.50 107.53
Total 5.69 128.68 93.00 227.37 177.90 0.821 107.50 286.21
The nitrogen balance in the paddy field can be divided into input, internal and output systems.
Chemical fertilizer, nitrogen fixation, precipitation and irrigation water are inputs. The nitrogen
mineralization in the soil is an internal factor. The inflow of total nitrogen was 221.7kg/ha (fertilizer
40.8%) and discharge was 286.2kg/ha (rice uptake 37.6%, surface runoff 62.2%). Total nitrogen
inflow was 1.79 ~ 0.00 kg/ha by precipitation, 30.31 ~ 0.01 g/ha by irrigation and 93.00 kg/ha by
fertilizer. Most of the nitrogen inflow was by irrigation and fertilization, and the amount of
precipitation was very small and can be ignored. The total nitrogen outflow was 286.21 kg/ha. And the
total nitrogen outflow by surface runoff was 58.60 ~ 0.03 kg/ha, and by infiltration 0.18 ~ 0.00 kg/ha.
The amount of rice uptake during the cultivation period was 107.5 kg/ha. Overall, the nitrogen outflow
by surface runoff and the amount of infiltration were very small and can be ignored.
Table 3 Nutrient balance (TP)

Input (kg/ha) Output (kg/ha)
Time
Rainfall Irrigation Fertilizer Subtatoal Runoff Infiltration Rice uptake Subtatoal
6/M 0.000 3.47 48.00 51.47 4.58 0.04 4.62
6/L 0.002 1.37 1.37 3.85 0.04 3.88
7/F 0.002 1.18 1.19 0.55 0.01 0.56
7/M 0.002 3.26 3.27 0.28 0.01 0.28
7/L 0.002 0.95 0.95 1.15 0.01 1.16
8/F 0.001 0.08 0.08 0.13 0.01 0.14
8/M 0.001 0.11 0.11 0.34 0.01 0.35
8/L 0.001 0.54 0.55 0.47 0.01 0.48
9/F 0.001 1.54 1.54 1.13 0.01 1.13
9/M 0.001 0.96 0.96 0.71 0.01 0.72
9/L 0.001 1.24 1.24 1.11 0.01 1.11
10/F 0.000 0.00 0.00 0.00 0.00 18.10 18.10
Total 0.014 14.70 48.00 62.73 14.31 0.17 18.10 32.53
The amounts of phosphorus added were 48.00 kg/ha (76.5%) from chemical fertilizer, 0.14 kg
/ha (0.02%) from precipitation and 14.7 kg/ha (23.4%) from irrigation water (Table 3). The output
loads were 14.31 kg/ha (44.0%) in drainage water and 18.1 kg/ha in rice grains (55.6%). The output
was much smaller than the input. About 56% of the phosphorus fertilizer was absorbed by paddy field
soil.
Table 4 Pollutant loads in paddy filed.

Water Input (kg/ha) Output (kg/ha) Net ouflow
Quality Rainfall Irrigation Subtatoal Runoff Infiltration Subtatoal loads (kg/ha)
BOD 15.4 157.4 172.8 178.8 0.9 179.7 6.8
COD 45.2 934.8 980.0 1001.0 5.1 1006.1 26.1
TOC 12.7 218.4 231.1 236.0 1.1 237.1 5.9
TN 5.7 221.7 227.4 177.9 0.8 178.7 -48.7
TP 0.01 62.7 62.7 14.3 0.2 14.5 -48.2
SS 11.8 2737.6 2749.4 1582.9 10.1 1593.0 -1156.3
The net outflow load of nutrients (the outflow load minus the inflow load) is one of the most
important factors in evaluating the role of paddy fields in water conservation. A negative value for the
net outflow load means that the catchment is a nutrient sink, and a positive value means it is a nutrient
source. Table 4 shows the input and output load. The total input load for BOD, COD, TOC, total
nitrogen, total phosphate and SS was 172.8 kg/ha, 980.0 kg/ha, 231.1 kg/ha, 227.4 kg/ha, 62.7 kg/ha
and 2749.4 kg/ha, respectively. And the total output load for BOD, COD, TOC, total nitrogen, total
phosphate and SS was 179.7 kg/ha, 1006.1 kg/ha, 237.1 kg/ha, 178.7 kg/ha, 14.5 kg/ha and 1593.0
kg/ha, respectively.
4. Conclusion
In Korea since the 1960s, the reclamation project has steadily progressed and the reclaimed area
has increased significantly. The reclamation project that was conducted earlier in an area where there
was a lot of food production for the purpose of financing is now used as agricultural land. At this time
many agricultural reservoirs and dikes were constructed. The major use of these estuarine reservoirs
was for irrigation of agricultural lands, but many estuarine reservoirs have water quality problems, and
many efforts have been undertaken to reduce point source pollution directly discharged to water
bodies. Non-point source pollutants are often significant contributors to eutrophication in lakes,
reservoirs and estuaries. Paddy fields are the main non-point source. Agricultural land is used for a
variety of farming methods depending on soil, local features, climate, irrigation and fertilization, and
the agricultural non-point pollutant source overflow rate largely depends upon the rainfall amount as
well as various agricultural activities. Therefore, an investigation was required of rice paddy overflow,
considering weather, farming, water quality management, and changes in the environment. Even
though there would be difficulty in quantification that is reproducible because of the complexity of the
characteristics of agricultural non-point pollutant sources, a close investigation and monitoring of the
loading characteristics of pollutant materials overflowing from rice paddies are needed. In this study, a
freshwater irrigated paddy field was evaluated for water quality, pollutant concentration and pollutant
loads in agricultural during the irrigation period in Korea in 2011. And water balance and nutrient
balance removal performance for water purification were analyzed. In 2011, the total influent of the
paddy field was 5256.8mm and effluent was 678.5mm. The water loss from the runoff was 4833.6 mm
(94.33 %) and evapotranspiration for the growth of plants was 260.1mm (5.08 %). Drainage discharge
increases or decreases in proportion to rainfall or irrigation. Beginning in mid-August, irrigation water
decreased because irrigation was reduced by the midsummer drainage. Total influent of total nitrogen
was 221.70 kg/ha, and effluent was 286.21 kg/ha. And total phosphorus influent was 62.73 kg/ha and
effluent 32.53 kg/ha. Most nutrients from nitrogen and phosphorus were supplied from fertilization
(40.90 % and 76.52 %, respectively). And nitrogen and phosphorus were effluents from rice uptake
(37.56 % and 55.64 %, respectively) and runoff (62.16 % and 43.96 %, respectively). All pollutants
under the influence of fertilizer in June are the largest. The discharge load was large in fertilization
and rice transplanting. The net outflow loads for BOD, COD, TOC, TN, TP, SS in the paddy field was
6.8 kg/ha, 26.1 kg/ha, 5.9 kg/ha, -48.7 kg/ha, -48.2 kg/ha and -1156.3 kg/ha, respectively. Therefore,
the study field was discharge type paddy of BOD, COD and TOC and absorption type paddy of TN,
TP and SS. Analysis of the results of the net discharge load of the paddy field found that rather than
pollutant sources of discharge in the range of general rainfall, there was a purification function from
the water conservation. Since monitoring of this study was performed for only five months, it was too
difficult to determine the general characteristics of non-point sources in reclaimed land. Therefore, it is
important to establish a non-point source database that has a high confidence level and various
research related parameters through continuous monitoring.
Acknowledgement
This work was supported by the ‘Ag-BMPs development project for water quality improvement in
Saemangeum estuarine reservoir’ funded by the Ministry of Agriculture, Food and Rural Affairs(MAFRA).
References
Ahn, J. H., Yun, S. L. and Kim S. K., 2012, “Runoff characteristics of non-point source according to
cultivation activity in in river district.” KSEE, Journal of Korea Society of Environ.Engineers,
34(7), 480-487.
APHA., 2005, “Standard methods for examination of water and wastewater, 21st edition. Washington:
American public health association.
Han, K. W., Cho, H. Y. and Choi, J. K., 1999, “Annual runoff loading of nitrogen and phosphorus
from a paddy field.” KSAE, Magazine of the Korean Society of Agricultural Engineers, 24(1),
29-33.
Jeon, J. H., Yoon, C. G., Hwang, H. S. and Yoon. G. S., 2003, “Water quality model development for
loading estimates from paddy field.” The Korean Society of Limnology, 36(3), 344-355.
Jeon, J. H., Yoon, C. G., Choi, J. G. and Yoon, G. S., 2005, “The comparison of water buget and
nutrient loading from paddy field according to the irrigation methods” The Korean Society of
Limnology, 38(1), 188-127.
Jung, J. W., Yoon, K. S. and Choe, W. J., 2008, “Improvement measures of pollutants unit-loads
estimation for paddy fields.” KSWQ, Journal of Korea Society on Water Environment, 24(3),
291-296.
Yoon, C. K., Hwang, H. S., Jeon, J. H. and Ham, S. H., 2003, “Analysis of nutrients balance during
paddy rice cultivation,” KSL, Korea Journal of Ecology and Environment, 38(1), 66-73.
Song, J. H., Kang, M. S., Song, I. H. and Jang J. R., 2012, “Comparing farming methods in pollutant
runoff loads form paddy fields using the CREAMS-PADDY model.”, KSAE, Magazine of the
Korean Society of Agricultural Engineers, 31(4), 318-327.
Y. W. Feng, I. Yoshinaga, E. Shirarani, T. Hitomi and H. Hasebe., 2003, “Nutrient balance in a paddy
field with a recycling irrigation system.” Diffuse Pollution Confernce.
THE FUTURE
F-15
Evaluation of Effects on Baseflow of Using Measured Field Slope

Length and Slope using SWAT
Ji Min Lee, Younghun Jung, Gwan Jae Lee, Seong Joon Kim,
Joong Dae Choi, Kyoung Jae Lim
ABSTRACT
Recently, localized rain and snow are happening all over the world due to dramatic climate changes.
In Korea, damage of human life is triggered by this. To solve this problem, this research used SWAT
model to predict outflow of small-and-large outlets, many types of soil and land use during long time.
However SWAT model calculates every outlets with same slope and slope length. This can reveal
problem in analysis of outlet properties. For this problem, this research applied measured slope and
slope length and result was NSE=0.74, R2= 0.84. This result was different from existed one (NSE
=0.63 R2= 0.79) and prove that this method can’t calculate outflow exactly. To be more accurately,
slope length that is less than 0.15 degree of slope applied as 400m which is average of measurements
showed result as NSE =0.73 R2= 0.83. Therefore, the average of slope length estimated from
individual sub-watersheds can be used to improve runoff estimation in the regions unavailable the
observed slope map and slope length. Accordingly, consideration of the slope map and slope length
will give a significant contribution to the hydrologic and hydraulic assessment by improving the
prediction ability of the existing SWAT model
1. Introduction
Recently there is sudden climate changes and natural disaster such as localized heavy rain and
heavy snow. Especially according to IPCC's 4th report, it predicts that there would be frequent natural
disaster due to climate changes. Korea also has huge number of casualties because of natural disaster.
To solve this problem, many environmental organization and governments try to manage water and
researchers carry out hydraulic/ hydrologic computer modeling. For analysis with computer, variety of
models are developed. Such as SWAT, HSPF, SWMM. Among them, SWAT is the most popular
around the world. SWAT, basins-unit model, is able to predict outflow and sediment at huge and
complex area reflecting variety of solid, land use and management in long-term period.
However, the SWAT model was developed based on the U.S. topography with typically mild
slope. From this reason, when the SWAT model is applied to the steep sloped region such as Korea,
some errors can occur in extracting topographic factor.
In particular, the SWAT model estimates the mean slope length of sub-watershed by using the
relationship between mean slope of sub-watershed and slope length of HRU. In such a process, the
sub-watersheds with steep topographic slope more than 25 % have a short uniform slope length of
0.05 m, which is not a physically meaningful number
Accordingly, it needs the application of the SWAT model considering topographic
characteristics of sub-watershed for accurate hydrologic assessment in sub-watershed.
The objectives of this study are 1) to predict runoff by applying the observed slope map and
slop length to the SWAT model and compare it with the existing SWAT model; and 2) using the
observed slope map and slope length; and 2) to suggest a direction for improvement in application of
the SWAT model when the observed data is not available. The objectives are implemented in
Haean-myeon watershed with the observed slope map and slope length.
2. Material and methods
2.1 Study area

The study area is a watershed of 63.08 km2 located in Haean-myeon, Yanggu-gun,
Gangwon-do, Republic of Korea (Fig. 1). Land use in the selected watershed consists of forest of
54.70 %, agriculture of 30.52 %, grassland of 9.85 %, residential and urban area of 4.88 %, and
water of 0.05 %. The watershed is surrounded by high mountains more than 1,000 m. For this reason,
the watershed is geologically isolated from other places in Yanggu-gun. Also, the water body in the
watershed of Haean-myeon is a tributary of SoYang River, as a part of Han River, but its evolution is
rarely developed. Its stream which is mainly comprised of direct runoff joins the Mandea stream, a
main stream in the watershed. The Haean-myeon watershed has geometric shape of an ellipse and
geological depression. So, this watershed is called as ‘Punch Bowl’ because its center is deeply carved
in. The elevation distribution in the watershed ranges from 400 m to 1,304 m. In such a regard, the
cross-section the watershed is U-shaped. From this reason, the average slope of the river in the
watershed is about 11°, and the river slope is gradually decreased from upstream (20°) to downstream
(5°).
Fig. 1 Location of the Haean-myeon watershed, Gangwon province, Korea
2.2 Introduction of SWAT and SWAT-CUP
SWAT is a semi-distribution rainfall-runoff model to reflect the changes in climate and land
use (Arnold et al., 1993), and has been widely used as one of poplar rainfall-runoff model (Ryu et al.,
2011). However, SWAT involves a lot of parameters associated with hydrologic processes such as
rainfall-runoff and is necessary to optimize them for accuracy of prediction. For calibration of the
model, SWAT provides Auto-Calibration tools with ARASOL and SUGLASSES techniques. Recently,
SWAT-CUP was developed by Swiss Federal Institute of Aquatic Science and Technology (EAWAG)
in order to incorporate various Web-based programs for automating calibration or uncertainty analysis
into SWAT. The program can run SUFI2 , GLUE, ParaSol , MCMC, and PSO in Window mode.
2.3 Constructing SWAT input data
In this study, DEM (Digital Elevation Model) was constructed using digital GIS maps
(1:5,000) provided by NGII (National Geographic Information Institute, Republic of Korea), and the
reconnaissance soil map (1:50,000) provided by the RDA (Rural Development Administration,
Republic of Korea) was used as a base soil map. Soil types in Haean-myeon watershed mainly consists
of Re (Silt-Sand), Rock (Silt-Sand), Ra (Silt-Sand), Mu (Clay-Slit), Ma (Silt-Sand), An (Clay-Slit).
2.4 Correction of outflow

Also, selecting an appropriate calibration algorithm according to their needs, users can
calibrate the SWAT model for the observations of multiple outlets. In addition, the periods of
observations used for calibration can be selected by users (e.g. periods of recession or flood) in whole
period. In this study calibration was performed by the SUFI-2 algorithm which has been used to
optimize the parameters of SWAT in many researches. Uncertainty in the calibration and prediction of
SWAT is typically represented by p-factor which is the percentage of observations covered by the 95
PPU (95 Percentage Prediction Uncertainty) and by r-factor which is the relative width of the 95PPU
band divided by the standard deviation. Here, individual a p-factor of 1 and r-factor of 0 represents a
simulation that perfectly corresponds to the measured data. Accordingly, the most sensitive parameters
of the SWAT model using SWAT-CUP were calibrated for 25 observed daily streamflow between
January 2010 and August 2010 measured at Moolgolkyo in the Hean-myeon watershed. The
calibrations were conducted for scenarios in Table 1.
n n
NSE = 1 − ∑ (Oi − Pi ) 2 / ∑ (Oi − Oi ) 2 (1)
i =1 i =1
O
Where, i denoted the observed runoff, and
P
i is the simulated runoff. O
i is the mean of the
observations. High NSE value means goodness of fit between the observations and the simulations,
and a NSE value of 1 indicates a perfect congruence. Ramanarayana et al. (1997) suggested that the
simulations with R2 more than 0.5 and NSE more than 0.4 reflect the observations well.
Table 1 Scenario1 vs. Scenario2 vs. Scenario3 for flow estimation

2010 (year)
Scenario1 SWAT model applied measured Slope length & Slope
Scenario2 Default SWAT model
Scenario3 slope length of 400 m to regions with slope lower than 0.15 degree
3. Result
3.1 Result for correction of outflow

In this study, the calibration of SWAT for runoff observed at Moolgolgyo in the Hean-myeon
watershed is conducted by adjusting several factors that is sensitive to the model output. The SWAT
calibration using observed slope map and slope length (scenario 1) resulted in NSE =0.74 and R2 =
0.84 (Fig. 2). In calibration of the existing SWAT (scenario 2) NSE and R2 were estimated as 0.63 and
0.79, respectively (Fig. 3). SWAT calibration applying slope length of 400 m to regions with slope
lower than 0.15 degree (scenario 3) produced NSE of 0.73 and R2= 0.83 (Fig. 4). Here, the average of
slope length for slope under 0.15 degree) is 400 m. From these results, application of scenario 3 to
SWAT calibration lead to less difference between the observations and the simulations than scenario 2
Table 2 Nine parameters used in calibration for flow estimation

Parameter Description Variation Method Value
CN2 USLE cropping and management (C) factor Multiply by Value -23.125
SURLAG Direct runoff lag time Replace by value 0.075
LAT_TIME Lateral flow travel time(days) Replace by value 1.4675
ALPHA_BF Baseflow alpha factor Replace by value 1.00
GW_DELAY Groundwater delay Add to value 9.9175
Threshold depth of water in the shallow

GWQMN Add to value 1.215
aquifer required for return flow to occur
GW_REVAP Groundwater ”revap” coefficient Add to value 0.0495
SOL_AWC Available water capacity of the soil layer Multiply by Value -0.0695
SOL_K Saturated hydraulic conductivity Multiply by Value 0.2
Fig. 2 SWAT model applied measured Slope length
Fig. 3 Default SWAT model
Fig. 4 slope length of 400 m to regions with slope lower than 0.15 degree
4. Conclusion
Application of the existing SWAT led to worse runoff estimation than the case considering the
observed slope map and slope length due to its inherent structural limitations. In this regard, baseflow
is also overestimated. Therefore, it is necessary to consider topographic characteristics such as slope
map and slope length in SWAT simulation. Also, the average of slope length estimated from
individual sub-watersheds can be used to improve runoff and baseflow estimation in the regions
unavailable the observed slope map and slope length. Runoff estimation using this method is also
sensitively affected by land use change, climate change, and crop distribution, but shows better results
than using the existing SWAT model. Accordingly, consideration of the slope map and slope length
will give a significant contribution to the hydrologic and hydraulic assessment by improving the
prediction ability of the existing SWAT model
References
Arnold JG, 1992, Spatial Scale Variability in Model Development and Parameterization, Ph.D.
Dissertation, Purdue University, West Lafayette, IN. 1-186
Ryu JC, Choi JW, Kang HW, Kum DH, Shin DS, Lee KH, Jeong GC, Lim KJ, 2010, Evaluation of
groundwater recharge for land uses at Mandae Stream watershed using SWAT HRU Mapping
module. Journal of Korea Society on Water Environment 28(5): 743-753
Ramanarayanan TS, Williams JR, Dugas WA, Hauck LM, McRarland AMS, 1997, Using APEX to
Identify Alternative Practices for Animal Waste Management. ASAE Inernational Meeting
97-2209
THE FUTURE
F-16
Assessment of Paddy Field Runoff on Water Quality of Yeongsan

River Basin by Load Duration Curve
Dongho Choi*, Jaewoon Jung*, Kwangsik Yoon*, Woojung Choi*, Hana Park**
*Dept. of Rural & Bio-systems Engineering, Chonnam National University, Korea
**Yeongsan River Environment Research Center, Korea
ABSTRACT
Paddy field could be source of pollutant or not depending on target water quality of ambient water
body. The objectives of this study were to characterize pollutant load from paddy fields by generating
load duration curve (LDC) and to identify exceedence frequency of load above total maximum daily
load (TMDL). LDC for paddy fields reflecting current weather conditions and agricultural practices
was generated through intensive monitoring from paddy fields located within Yeongsan River basin in
Korea. Target concentrations of BOD and T-P, which were designated by TMDL program of
Yeongsan River basin were multiplied by runoff volume of paddy fields to develop target load duration
curve. Exceedence frequency above TMDL was evaluated by comparing target and observed LDCs.
This study demonstrated that LDC method is feasible to evaluate how often paddy field behaves as
pollutant source under specific water quality goal.
KeyWords: LDC, paddy field, TMDL, water quality
1. Introduction
Water quality management for streams and lakes in South Korea has been implemented on point
sources such as industrial and domestic sewage waste. In spite of such efforts, however, the water
quality of streamsand lakes has not been improved satisfactory. Pollutant load by drainage from
agricultural non-point sources such as pesticides and fertilizer are suspected for deterioration of water
quality in streams and lakes. Recently, water quality of drainage from agricultural land are getting paid
attention due to Total Maximum Daily Load(TMDL) program in Korea, where paddy fields occupy
more than 60% of the total farmland. To maintain or improve river and lake water quality, a
quantitative estimation of the pollutant loads from the non-point sourcesis necessary. Therefore,
several studies have been conducted to determine nutrient losses by surface drainage from paddy fields
in Korea, and various results have been reported (Cho et al., 2002; Yoon et al., 2003; Cho et al., 2003;
Yoon et al., 2006). However, studies regarding reasonable total maximum daily load from paddy
fields are few.
The load duration curve (LDC) method recommendedby the EPA (USEPA, 2007) provides a
simpler way to estimate watershed loading in TMDL development. Several researchers (Stiles, 2001;
Cleland, 2002, 2003; Bonta and Cleland, 2003; and O’Donnell et al., 2005) have utilized LDCs to
estimate the TMDL. The major steps of this method include: (1) generating a flow duration
(cumulative frequency) curve based on available historical flow data; (2) calculating TMDL by
multiplying the numerical water quality target with flows, and plotting against the cumulative
frequency; (3) estimating the existing instantaneous loads by multiplying the observed ambient water
quality data with the averaged flows on the sampling dates, and plotting against the corresponding
cumulative frequency on the LDC; and (4) calculating the margin of safety, load allocation and
reduction for different flow regimes based on the LDC. This method is simple and straightforward,
and particularly useful in addressing the essential role played by flows in determining load capacity
(Shen and Zhao, 2010).
Load duration curves are relatively easy to develop once one has an understanding of how they
work. Most resource management personnel with a background in hydrology and water quality should
be able to develop and interpret load duration curves with relatively little training. Similarly,
explaining the results of a load duration curve to the public can be easier than explaining other
technical approaches, such as modeling. This can promote effective communication between TMDL
developers and those responsible for implementation (Cleland, 2002). The objectives of this study
were to characterize pollutant load from paddy fields by generating LDC reflecting current weather
conditions and agricultural practices through intensive monitoring on paddy fields in Yeongsan River
basin and to identify load exceedence frequency from paddy field above TMDL to meet target water
quality designated for Yeongsan River.
2.1. Description of paddy fields

The study was conducted during a period of five crop-years (from 2008 to 2012) in a rice
cultivation area located in Emda-myun, Hampyeong gun, Jeollanam-do, Korea (35°02 ′ 11″N,
126°31′29″E), (Fig. 1). The studied paddy field area was 13.69 ha and it was composed of several
paddy plots (100 m by 50 m) separated by irrigation and drainage canals. Average annual precipitation
ranged from 1007.2 mm to 1626.8 mm, and more than 60~80% of this occurred during the cropping
period (May-September). The physical soil properties of the experimental paddy field was of
Pyeongtaeg series (silt loam, mixed, mesic family of Typic Endoaquepts) (National Institute of
Agricultural Science and Technology, 2008).
The crop period was from the beginning of May to the end of September. Rice was transplanted
from May 26 to June 5 and basal fertilizer was applied between late May and early June. Additional
fertilization (tillering and panicle fertilization) were applied late June and early August. Application rate of
fertilizer to paddy fields was estimated after personal, face to face interviews with farmers. Applied basal
fertilizer rates was 55.5 kg N ha-1, 51.5 kg P2O5 ha-1, 51.5 kg K2O ha-1, tillering fertilizer was applied at a
rate of 27.2 kg N ha-1 and panicle fertilizer was applied at a rate of 46.5 kg N ha-1.
Fig. 1 Schematic representation of study area and location of sampling stations
2.2. Precipitation and flow measurements

In the study field, sets of rain gauges (Casella Rainfall System, UK), five sets of water level
gauges (OTT model Orphimedes, Germany), an infiltrometer, an evaporation pan were installed. The
rain gauge was set up house near the experimental rice paddy field for automatic recording. Other
meteorological data (temperature, windspeed, relative humidity etc.) were obtained from the Rural
Development Administration (RDA). The water level in irrigation canal was measured using a water
level gauge with data logger. The irrigation amount was decided by subtracting flow at the endpoint of
canal from the flow of intake. A water level gauge was set up at the outlet of the drainage canal to
measure discharge from the paddy fields. Flow velocity in the irrigation and drainage canal was
measured by a flow velocity meter (OTT model C2, Germany) and flow rates were calculated using
mid-section method. The measured water levels were converted to flow rates using a water level-flow
rate relationship derived from the measured data.
2.3. Samplings and chemical analysis

Samples of precipitation and irrigation water were taken for each event. During storm periods,
drainage water was sampled, by auto-sampler (ISCO sampler 6712, USA), several times proportion to
discharge amount at the outlet of the main drainage canal whenever an event occurred. Bottles
containing sample were placed in insulated chests, and crushed ice was then added to fill the chests.
All samples taken for chemical analysis were refrigerated at 0 to 4°C soon after collection until
analysis. Concentrations of five constituents –chemical oxygen demand (COD), total nitrogen (T-N),
total phosphorus (T-P), and suspended solid (SS), total organic carbon (TOC) – were analyzed
according to APHA methods (2001).
2.4 Development of FDC (Flow Duration Curve) and LDC(Load Duration Curve)
A duration curve is a graph representing the percentage of time during which the value of a given
parameter (e.g. flow, load) is equaled or exceeded. Generally, the percentage of timeduring which
specified flows are equaled or exceeded may be compiled in the form of a flow durationcurve. This is
a cumulative frequency curve of daily mean flows without regard to chronology ofoccurrence
(Leopold, 1994). Using available daily runoff data, a flow duration curve was developed for the study
site. Data for the curve is generated by: 1) ranking the daily runoff data of paddy field from highest to
lowest; 2) calculating percent of days these flows were exceeded. The load duration curve is
developed by calculating daily loads for each sample using the pollutant concentration and discharge
for the particular day.Next, the flow values for each day are compared to the flow duration curve data
in order to determine the value for ‘percent of days flow exceeded’ which is equivalent to ‘percent
of days load exceeded’. (Shen and Zhao, 2010)
3.1 Observed rainfall, irrigation, and runoff

Observed rainfall, irrigation and runoff from the paddy field during cropping period is shown in
Fig 2. Annual rainfall amount ranged 1007.2~1626.8 mm and 678.0~1160.0mm (avg. 891.6 mm)
occurred during cropping period, which was about 70% of annual rainfall. Rainfall amount during
non-cropping period (Oct. – Apr.) was 329.2~466.8 mm and evenly distributed. Irrigation amount
ranged 599.9~1091.5mm (avg. 806.6 mm) during cropping period. Runoff amount ranged
503.8~837.0mm (avg. 704.3 mm) during cropping period each year.
Fig. 2 Observed rainfall, irrigation, and runoff from the paddy field .
3.2 Observed concentrations of BOD, COD, TOC, T-N, and T-P from the paddy fields.
Total 105 storms were sampled by auto-sampler and event mean concentrations were
determined and 45 samples were collected for non-storm period by grab technique. Observed
concentrations during storm period from the paddy field were 0.05~10.86(avg: 3.98) mg/L for BOD,
2.26~22.32(avg: 10.96) mg/L for COD, 2.58~25.77(avg: 8.69) mg/L for TOC, 0.06~27.36(avg: 3.18)
mg/L for T-N, 0.02~1.96(avg: 0.43) mg/L for T-P, and 1.25~1038.75(avg: 95.12)for SS mg/L,
respectively. Observed concentrations during non-storm period from the paddy field were
0.05~6.70(3.56) mg/L for BOD, 6.64~19.18(12.99) mg/L for COD, 2.93~58.75(11.04) mg/L for TOC,
0.34~18.98(3.60) mg/L for T-N, 0.05~1.14(0.30) mg/L for T-P, and 1.25~698.0(76.90) for SS mg/L,
respectively.
Fig. 3 Observed concentrations of BOD, COD, TOC, T-N, and T-P from the paddy fields.
3.3 Flow Duration Curve (FDC)

Based on hydrologic conditions, the flow duration curve can also be divided into different flow
zones or intervals. It consists of ordered flows classified into high flows (0–10% exceedence), moist
conditions(10–40% exceedence), mid-range flows (40–60% exceedence), dry conditions (60–90%
exceedence), and low flows (90–100%exceedence). Figure 4 presents a flow duration curve using data
from the paddy field within Yeongsan River basin. A total 748 days of discharge was observed, which
was equivalent of 40.9% of study period (1827 days). About 60% of discharge from paddy field was 0
mm/day since there was no irrigation and only 30% of annual rainfall occurring during non-cropping
period. The figure 4 illustrates that the highest observed flow value at this gage for the period of
record is 99.38mm/day and the lowest observed flow is 0 mm/day. The discharge equivalent 10, 20, 30,
and 40% exceedence were, 4.81, 2.38, 1.39, and 0.07 mm/day, respectively.
Fig. 4 Flow duration curve of the paddy fields during the study period.
3.4 Load Duration Curve (LDC)
The LDCs provide the pollutant load at the monitored location and under different hydrologic
conditions. These curves can be used as a reference for guiding pollutant load reduction efforts in the
watershed. Load duration limit curve can be created from a flow duration curve by multiplying the
flow values by the applicable water quality criterion or target. The limit curve therefore represents the
allowable load (or the TMDL) at each flow condition. The observed loads, which are calculated by
multiplying the sampled pollutant concentration by the instantaneous flow associated with the sample.
Points plotting above the curve represent exceedences of the target and are therefore unallowable loads.
Those plotting below the curve represent compliance with the target and allowable daily loads.
Figure 5 displays the LDC of the COD, TOC, T-N, and SS with monitored data. BOD and T-P
are currently only concern of TMDL program in Korea. Target concentrations of BOD and T-P are
designated at the outlet of the seven subwatersheds in Yeongsan River basin in Korea. Since
non-storm period water quality measured weekly or biweekly, load of unsampled day was estimated
by linear interpolation of measured concentrations. The estimated daily load was 0.00~88.02(1.10)
kg/day for BOD, 0.00~167.21(2.83) kg/day for COD, 0.00~144.49(2.40) kg/day for TOC, 0.00~47.10(0.78)
kg/day for T-N, 0.00~13.88(0.08) kg/day for T-P, and 0.00~1675.83(16.19) kg/day for SS, repectively.
Fig 6 show estimated LDC from paddy field and target LDC generated by target concentration
and FDC of paddy field. To investigate exceedence frequency of load from paddy field above water
quality goal of several stations in Yeongsan River basin, several target load duration curve were
generated for BOD, COD, TOC, T-N, and T-P, and points plotting above the curve representing
exceedences of the target were counted to evaluate unallowable load occurrence (Table 1). The results
show that paddy field exceed BOD goal 8.2~11.6% where target concentration was set as
5.2~5.6mg/L (station YB-B, YB-C, YB-D), but exceeded upto 30.7~37.7% where target concentration
was set 2.1-2.7 mg/L (station YB-A, YB-E, HR-A, JS-A). In the same manner, paddy field exceed T-P
goal 3.2~8.0% where target concentration was set as 0.35-0.62mg/L (station YB-B, YB-C, YB-D), but
exceed 23.0~30.7% where target concentration was set as 0.13~0.17mg/L (station YB-A, YB-E,
HR-A, JS-A).
200 200
150 150
COD (kg/day)
TOC (kg/day)
100 100
50 50
0 0
0 20 40 60 80 100 0 20 40 60 80 100
Percent of Days Flow Exceeded (%) Percent of Days Flow Exceeded (%)
150 1000
120 800
90 600
T-N (kg/day)
SS(kg/day)
60 400
30 200
0 0
0 20 40 60 80 100 0 20 40 60 80 100
Percent of Days Flow Exceeded(%) Percent of Days Exceeded(%)
Fig. 5 Observed LDC of COD, TOC, T-N, and SS of the paddy fields.
200 20
150 Target BOD 2.1 mg/L 15 Target T-P 0.151 mg/L

Target BOD 5.6 mg/L Target T-P 0.620 mg/L
BOD (kg/day) 100 Observed BOD 10 Observed T-P
T-P (kg/day)
50 5
0 0
0 20 40 60 80 100 0 20 40 60 80 100
Persent of Flow Exceeded(%) Persent of Flow Exceeded (%)
Fig. 6 Estimated and target LDC of BOD and T-N of the paddy fields.
Table 1 Target water quality of TMDL stations in Yeongsans River basin and exceedence probability of
load from paddy field above target load .
BOD T-P
Target Exceedence Target Exceedence
Station
concentration Probability (%) concentration Probability (%)
(mg/L) (mg/L)
YB-A 2.1 37.7 0.151 23.4
YB-B 5.6 8.2 0.620 3.2
YB-C 5.2 11.6 0.428 4.8
YB-D 5.2 11.6 0.350 8.0
YB-E 2.4 23.0
HR-A 2.2 34.4 0.130 23.9
JS-A 2.7 30.7 0.171 20.3
4. Summary and conclusions
Paddy field could be source or sink of pollutant depending on target water quality ofambient
water body. To evaluate how often paddy field behaves as pollutant source, LDC was developed using
5-year monitored data from paddy field in southern Korea. Exceedence frequency as pollutant source
was evaluated by comparing target LDC of water quality specified by Yenongsan River basin TMDL
program. Study found that the maximum exccedence frequency of the paddy fields must be below 40~
50% since paddy runoff does not generated from paddy field during late fall to early spring while
stream maintain low flow. The LDC method showed that the paddy fields only exceed BOD goal
8.2~11.6% where target concentration was set as 5.2~5.6mg/L, but exceeded upto 30.7~37.7% where
target concentration was set 2.1-2.7 mg/L. Meanwhile, the paddy fields exceed T-P goal 3.2~8.0%
where target concentration was set as 0.35-0.62mg/L, but exceed 23.0~30.7% where target
concentration was set as 0.13~0.17mg/L. This study demonstrated that LDC method is feasible to
evaluate how often paddy field behaves as pollutant source under specific water quality goal. However,
further study is required since sufficient flow data are needed to establish return frequencies, and a
significant amount of concentration data especially for non-storm period to compare to the limit curve.
Acknowledgements
This work was supported by ‘Basic research program of Yeongsan and Seomjin River system’
through the Yeongsan River Environment Research Center and the ‘Ag-BMPs development project
for water quality improvement in Saemangeum estuarine reservoir’ funded by the Ministry for Food,
Agriculture, Forestry and Fisheries (MIFAFF).
References
American Public Health Association, 2001, Standard methods for the examination of
waterand waste water, (21sted.). Washington, DC.
Cleland, B., 2002, TMDL development from the “bottom up” part II: using duration curves to connect
the pieces. In:Proceedings of National TMDL Science and PolicyConference. Water
Environment Federation, Phoenix, AZ,USA.
Cleland, B., 2003, TMDL development from the “bottom up”Part III: duration curves and wet-weather
assessments. In:Proceedings of National TMDL Science and PolicyConference. Water
Environment Federation, Chicago, IL,USA.
Leopold, L.B., 1994, A View of the River. Harvard University Press. Cambridge, MA.
Stiles, T.C., 2001, A simple method to define bacteria TMDLs inKansas. In: Proceedings of TMDL
Science Issues Conferences.Water Environment Federation and Association of State and
Interstate Water Pollution Control Administrators,Alexandria, VA and Washington, D.C., USA.
Bonta, J.V., Cleland, B., 2003, “Incorporating natural variability,uncertainty, and risk into water
quality evaluations usingduration curves” Journal of the American Water Resources
Association 39 (6): 1481-1496
O’Donnell, K.J., Tyler, D.F., Wu, T.S., 2005, TMDL report: fecal andtotal coliform TMDL for the
new river, (WBID 1442). In:Proceedings of the 3rd Conference of Watershed Management
to Meet Water Quality Standards and Emerging TMDL. Atlanta, GA, USA
US Environmental Protection Agency (US EPA), 2007, An Approachfor Using Load Duration Curves
in the Development of TMDLs. US EPA, Washington, DC, USA
Shen J. and Zhao Y., 2010, “Combined Bayesian statistics and load duration curve method for bacteria
nonpoint source loading estimation” Water Research 44:77-84.
National Institute of Agricultural Science and Technology, 2008, Taxanomical classification of Korean
Soils, NIAST, Suwon, Korea, pp. 464-465.
Yoon, Chun G., Ham J.H., Jeon J.H., 2003, “Mass balance analysis in Korean paddy rice
culture”Paddy and Water Environment 1(2):99-106.
Yoon, Kwang-Sik, Jae-Young Cho, Jin-Kyu Choi, and Jae-Gwon Son, 2006, “Water management and
N, P Losses from paddy fields in Southern Korea” Journal of the American Water Resources
Association (JAWRA) 42(5):1205-1216.
Cho Jae-Young, Kang-Wan Han and Jin-Kyu Choi, 2000, “Balance of nitrogen and phosphorus in a
paddy filed of central Korea” Soil Science and Plant Nutrition 46(2): 343-354.
Cho Jae-Young, Kang-Wan Han, Choi Jin-Kyu, Kim Young-Joo, and Yoon Kwang-Sik, 2002, “N
And P Losses from Paddy Field Plot in Central Korea” Soil Sci. Plant Nutr., 48(3):301-306.
Poster Session
THE FUTURE
P-03
An Analysis of Runoff Characteristics of Hosan Stream Using

Rainfall-Runoff Model
Seung J. Maeng*, Ji H. Shim*, Gil S. Hwang**, Dong O. Kim**, Ji H. Jeong*

*Department of Agricultural & Rural Engineering, Chungbuk National University, Cheongju,
Korea
**Halla E & C Research & Development Center, Seoul, Korea
In this study, the long-term runoff volume from the downstream part of Hosan Stream was estimated
using the SSARR (Stream Synthesis and Reservoir Regulation) model, a hydrologic model for
identifying the flow characteristics of the upstream part of a stream or river, due to the construction of
a comprehensive development complex for the Hosan Port breakwater and an LNG production plant
at the mouth of Hosan Stream. The parameters of the SSARR model should be estimated through
calibration with the measured runoff volume. As there are no measurement data from the selected
watershed, however, those of the Pyeongchang River watershed, which has similar physical
characteristics and reliable measurement data, were applied. In addition, the average daily runoff for
each year was estimated by simulating the daily runoff from 1972 to 2011.
In this regard, it is expected that measures to reduce the damages caused by floods can be devised
only through the analysis of consistently secured measurement data on the changes in rivers due to the
construction of an LNG production plant and a comprehensive development complex for the Hosan
Port breakwater.
Keywords:
1. Introduction
Today, irrigation facilities and facilities around rivers are being managed through their division
into low- and high-flow purposes, and they are being constructed in a way that can maximize their
economic effects. As a basic level of research and analysis prior to the installation of such facilities,
hydrologic and hydraulic analyses should be carried out. To date, hydrologic analysis has been
developed through many application.
In this study, the long-term runoff of the downstream part of Hosam Stream was estimated using
a hydrologic model, to identify the flow rate characteristics of the upstream part owing to the
construction of an LNG production plant and of a comprehensive development complex for the Hosan
Port breakwater in the downstream part of Hosan Stream.
2. Selection and Characteristics of the Model
This study was carried out by measuring the runoff volume of the Hosan Stream watershed, and
its calibration, via hydrologic analysis. To ensure the objectivity of the analysis results, a model was
selected among the numerous hydrologic models that have been introduced in South Korea and
overseas.
As long-term runoff analysis should be performed for hydrologic analysis, a long-term model is
required in terms of the period of analysis, and a lumped model that facilitates the acquisition of input
data rather than a distributed model that requires detailed information on rivers needs to be used as an
analysis method. Thus, the SSARR model was selected for this study as it is consistent with the
purpose of this study and has many national and international application cases.
Since its development by the U.S. Army Corps of Engineers (USACE) in 1956, the SSARR
model has constantly been calibrated, and it has been widely used for the purpose of reservoir
regulation and of the prediction of the real-time daily discharge of large-scale basins. This model,
which includes the watershed routing, hydraulic channel routing, and reservoir regulation models, has
been successfully applied to large rivers such as Columbia River in the United States and Mekong
River in Vietnam, and can estimate the runoff not only due to rainfall but also due to snowfall. In
addition, SSARR as a lumped-parameter model is capable of finding the optimal values for more than
24 parameters through a trial-and-error method. The time interval for calculation can range from 6
minutes to 24 hours, and several parameters, including SMI (soil moisture index), ETI
(evapotranspiration index), and BII (baseflow infiltration index), are given in the form of indicators
(USACE, 1991).
3. Features of the Selected Watershed
3.1 General conditions of the watershed

As a local river, Hosan Stream takes its rise from Icheon-ri, Wondeok-eup, Samcheok-si, flows
into the southeastern side, joins a small river at the point of Molgol in Icheon-ri, Wondeok-eup,
Samcheok-si, continues to flow into the southeastern side, and is then introduced into East Sea at
Hosan-ri during its flow, after confluence with Gilgok Stream, a local river in Ogwon-ri,
Wondeok-eup, Samcheok-si. Hosam Stream has a 20.5 km stream length and a 64.8 km2 watershed
area. It has a relatively gentle stream slope measuring 1/58-1/195 at the mid- and downstream parts,
but it maintains the form of a steep mountain stream with a stream slope of 1/17-1/28 at the mid- and
upstream parts.
Figure 1. Hosan Stream Watershed
3.2 Weather conditions and climate
To identify the weather conditions in the Hosan Stream watershed, the meteorological
observation scope of Uljin Regional Meteorological Office, which has geographical proximity to the
target watershed, was set as the survey area, and the characteristics by major meteorological element
were analyzed using the meteorological data for the last 13 years (1999-2011). Although the
observation scope of Samcheok-si belongs to Donghae Weather Station, the meteorological data of
Uljin Regional Meteorological Office were used because the survey area is located in the southern part
of Samcheok-si, thus having a smaller separation distance with Uljin Reigional Meteorological Office.
The results of the survey on the temperature data from Uljin Regional Meteorological Office for
the last 13 years (1999-2011) showed that the annual average temperature is 12.9°C and that the
seasonal temperature is 11.9°C in spring, 22.0°C in summer, 14.8°C in autumn, and 3.0°C in winter,
indicating a maximum temperature difference of 19.1°C. The monthly average temperature was found
to be highest in August (23.5°C), the monthly maximum temperature (36.3°C) also highest in August,
and the monthly minimum temperature (-13.3°C) highest in January.
For hydrologic analysis, it is important to secure the long-term rainfall data. From the rainfall
data collected by Uljin Regional Meteorological Office in the observation reference years 1972-2011,
the annual average precipitation was found to be 1,096.1 mm, and the seasonal rainfall distribution
was found to be 195.6 mm in spring, 485.6 mm in summer, 312.9 mm in autumn, and 102.1 mm in
winter, showing the highest precipitation in summer, which is the general phenomenon in the Korean
peninsula. The rainfall concentration ratio turned out to be 44.3% (Korea Meteorological
Administration, 2011).
4.1 Physical parameters

The parameters set in the hydrologic model can be largely divided into the physical,
hydro-meteorological, and internal processing parameters. The parameters to be determined in the
SSARR model include the hydro-meteorological input parameters, rainfall and snowfall data, index
weight, maximum amount of blocking, and temperature.
4.2 Internal processing parameters

The internal processing parameters of the SSARR model, such as the SMI-ROP (runoff percent),
BII-BFP (base flow percent), and S-SS (surface-subsurface separation), were estimated. The internal
processing parameters should be estimated through the calibration between the observed discharge
from Hosan Stream and the simulated discharge by the SSARR model. However, since there is no
observation discharge data from the water system selected in this study, calibration through the
comparison of the observation and simulation discharges using the SSARR model is realistically
impossible.
As such, in this study, the internal processing parameters were estimated by applying the
SSARR model on Pyeongchang River watershed, which is geographically close to the selected
watershed, has similar physical characteristics, and secures reliable observed discharge data as the
primary target, were used (Korea Water Resources Corporation, 2008). The internal processing
parameters, such as the SMI, ROP, BII, BFP, and S-SS, are shown in Table 1.
In addition, the values presented in the SSARR Manual were used for the internal processing
parameters, such as the infiltration storage period and the maximum infiltration and maximum
groundwater discharge ratios (USACE, 1991).
Table 1. Internal Processing Parameters
Surface Comp.
SMI (cm) ROP (%) BII (cm/d) BFP (%) S-SS (cm/hr)
(cm/hr)
0 8 0.0 40 0.0 0.00
1 30 1.0 14 0.5 0.15
2 51 1.5 11 1.0 0.57
3 65 2.0 10 1.5 1.07
4 81 2.5 10 2.0 1.57
5 59 3.0 10 2.5 2.07
10 100 5.0 10 3.0 2.57
999 100 10.0 10
4.3 Estimation of simulation discharge

Through the application of the determined input data and parameters, the daily runoff from the
end point of the Hosan Stream watershed from 1972 to 2011 was simulated using the SSARR model.
Figure 2 shows the schematized mean value of the daily runoff during the entire analysis period, and
those of the daily runoffs from 1972 to 1980, from 1981 to 1990, from 1991 to 2000, and from 2001 to
2011 with the Hosan Stream watershed.
Figure 2. Mean Value of Daily Runoff by Year
Of the daily runoff simulated for the past 40 years from 1972 to 2011, the largest runoff from
the Hosan Stream watershed occurred in 2003, the medium runoff in 1984, and the smallest runoff in
1979. In addition, the runoff after 2000 was found to be relatively greater compared to that before
2000.
4.4 Flow regime analysis

Using the simulation discharge of the Hosan Stream watershed estimated by using the SSARR
model, flow regime analysis, which can determine the flood discharge, abundant water discharge,
normal flow, low water discharge, and drought flow of the Hosan Stream watershed, was carried out.
As a result, it was found that the flood discharge is 89.5 ㎥/s, the abundant water discharge 0.8 ㎥/s,
the normal flow 0.4 ㎥/s, the low water discharge 0.3 ㎥/s, and the drought flow 0.2 ㎥/s.
5. Conclusion
In this study, an analysis on the runoff characteristics of the downstream part of Hosan Stream
was performed using the SSARR model, a long-term runoff model. Due to the absence of
measurement data from the selected watershed, the parameters of the Pyeongchang River watershed,
whose physical characteristics are similar to those of the selected watershed, and which has reliable
measurement data, were applied as the parameters for the model.
As a result of the estimation of the runoff with the SSARR model using the simulation discharge
of the selected Hosan Stream watershed from 1972 to 2011, the flood discharge was found to be 89.5
㎥/s, the abundant water discharge 0.8 ㎥/s, the normal flow 0.4 ㎥/s, the low water discharge 0.3
㎥/s, and the drought flow 0.2 ㎥/s.
It is expected that countermeasures against changes in rivers due to the construction of an LNG
production plant and a comprehensive development complex for the Hosan Port breakwater will be
devised by performing analysis through the consistently secured Hosan Stream watershed
measurement data in the future.
Acknowledgments
This research was supported by a grant (code: EW13-07-10) from the Center for Aquatic
Restoration of the Eco-STAR Project of the Korean Ministry of Environment.
References
Korea Meteorological Administration, 1999-2011, Annual Report of Weather Data.

Korea Water Resources Corporation, 2008, Report on the Establishment of the Han River Basin Real-Time
Reservoir Operating System (Analysis Model Development).
USACE, 1991, SSARR User Manual. North Pacific Div.
THE FUTURE
P-04
Development of Irrigation Management Method for Reducing

Inflow of Radioactive Substances in Japan
*Moono Shin , **Tomijiro Kubota, **Koji Hamada , **Tadayoshi Hitomi

*Agricultural Radiation Research Center, Tohoku Agricultural Research Center, National
Agriculture and Food Research Organization, 50 Harajuku-minami, Arai, Fukushima,
Fukushima 960-2156, Japan e-mail: moono@affrc.go.jp
**Department of Hydraulic Engineering, National Institute for Rural Engineering, National
Agriculture and Food Research Organization, 2-1-6 Kannondai, Ibaraki, Tsukuba 305-8609,
Japan. e-mail: tomi@affrc.go.jp
ABSTRACT
Radioactive contamination has been brought in the wide area by the Tokyo Electric Power Company
Fukushima Daiichi Nuclear Power Station disaster by the Tohoku earthquake and tsunami in Japan
on 11 March 2011. Decontamination work by the government or the local autonomy is going on after
the disaster. In paddy field in particular, a number of decontamination techniques have been
developed. However, considering long-term management to avoid re-contamination by radioactive
compounds again, and elucidating the transition of radioactive substances with agricultural water,
are required. Therefore, in this study, we introduced a system of turbidity monitoring in an irrigation
area of Fukushima prefecture to reduce the transition from radioactive substances that was flowed
into the upstream to canal or paddy field in downstream. Using relations of turbidity and
concentration of radioactive substances in the agricultural water, we developed a new irrigation
management method to reduce inflow of radioactive substances and analyzed the effect of the system.
1. Introduction
The wide area in Fukushima was contaminated with large amount radioactive substances by
Tokyo Electric Power Company Fukushima Daiichi Nuclear Power Station disaster by the Tohoku
earthquake and tsunami in Japan on 11 March 2011. In the aftermath, plantings are limited in some
areas and various decontamination works by government has continued a wide area though it reached
the third year after the earthquake.
In the paddy fields, many techniques of decontamination such as surface soil removing, upside
down plowing, and rotary mixing tillage with land preparation have been developed (Naka et al.,
2012).
To keep the effect of decontamination it should avoid recontamination. But the recontamination
with inflowing radioactive substances by agricultural water has been apprehensive (Kubota et al.,
2013).
Therefore, it is the important issues for revival after the disaster in the future, to analyze the
migration of radioactive substances with agricultural water, to reduce inflow of the deposited
radioactive substances from the environment to the paddy field or canal system.
If the radioactive substances in water are flowed into main canal, administrator can take some
actions for reducing inflow to paddy field. They are to stop the intake from the weir for stopping the
irrigation, to contact ditch rider to stop the division works for discharging agricultural water through
the drainage canal into the river, and to notice the information to farmers. In order to achieve the water
management such operations, it is necessary to measure in real time the concentration of radioactive
substances in the main canal (Figure 1).
Figure 1 Outline of irrigation management by system Figure 2 Study area in Fukushima, Japan
of reducing inflow of radioactive substances
Government, local government and the Universities have conducted the monitoring and
researches about radioactive substances in ponds and rivers. In the results, the radioactive cesium has
been detected from some turbid water during rainfall but hardly from the water during calm water
(Ministry of Agriculture, Forestry and Fisheries, 2013; Forestry and Forest Products Research
Institute, 2012).
We focused on the amount of radioactive substances detected from the turbid water during
rainfall. Then we introduced a turbidity monitoring system for a D irrigation system that is under the
management of an H land improvement district in Abukuma River basin, Fukushima.
The purpose of this study is to elucidate the relationship between concentration of radioactive
substances and turbidity of agricultural water in the D irrigation system and to examine the feasibility
of reducing the inflow of radioactive substances into paddy fields by turbidity.
2.1 Study area

The D irrigation system is located approximately 60km from the Tokyo Electric Power
Company Fukushima Daiichi Nuclear Power Station disaster, the H land improvement district has
about 1,900 members, 740ha of a beneficiary area and about 2.4 m³/s of water rights in the irrigation
period (Figure 2).
Tohoku Electric Power Co., which is operating the hydroelectric power plant in the Abukuma
River basin, has done operation the intake of the weir as a water source. The H land improvement
district determines the amount of water required for the water allocation and requests the amount of
water from the staff of Tohoku Electric Power Co.
In normal period there is no rainfall, the staff go check the weir after alternating 13:00 everyday,
perform the operation of intake at that time. Therefore, the administrator of the H land improvement
district can only request in before 13:00. But in the case of the weather warning, since the staffs of
Tohoku Electric Power wait all day long, the operation request of the weir can be accepted anytime.
2.2 Turbidity monitoring system

We have established two turbidity meters, an automatic water sampler, and five water gauges in
the D irrigation system (Figure 3).
In particular, the equipment in upstream point is able to browse homepages on the Internet to
use the FOMA (Freedom Of Mobile multimedia Access) cell line. Also administrator of the H land
improvement district is notified by e-mail when the turbidity crosses the standard turbidity.
Figure 3 The turbidity monitoring system and irrigation network
Table 1. The area of irrigation block
Total
Block 1 2 3 4 5 6 7 8 9 10
(ha)
Area
5.9 25.3 25.2 16.8 11.6 4.8 9.7 60.6 41.4 82.3 283.6
(ha)
Total
Block 11 12 13 14 15 16 17 18 19 20
(ha)
Area
83.0 26.7 59.0 80.8 18.1 11.8 88.9 15.2 62.6 10.7 456.8
(ha)
2.3 Analytical method of radioactivity
In order to verify the relationship between concentration of radioactive substances and turbidity
of agricultural water, we perform water sampling periodically in the irrigation period of 2013. The
subject of analysis is Cesium 134 and Cesium 137, the measurement was performed in germanium
semiconductor radiation detectors. For that radioactive material contained in agricultural water is often
low concentrations, some samples were concentrated by heating as a pretreatment. The accuracy was
measured at RSD (Relative Standard Deviation) 5% (eq. 1).
( : standard deviation, : mean) (1)
Figure 4 Radioactivity and turbidity in main canal Figure 5 Radiological dosage and turbidity in main
upstream canal middle
Table 2. Water management by H land improvement district in the irrigation period of 2013
Month
May June July August September Total
Discharge
Control
3 2 5 8 5 23
(time)
Stop
- - 7 4 7 18
(day)
Figures 4 and 5 show results to compare the turbidity and the concentration of radioactive
substances of 4 middle point and 14 upstream point samples of the main canal from May to August
2013, respectively. Higher the turbidity is, the higher the concentration of radioactive substances is.
The coefficient of determination n R2 is about 0.92 and 0.97 in upstream point and middle point,
respectively. It is possible to deduce the concentration of radioactive substances in a simple manner by
monitoring the turbidity of agricultural water.
Table 2 shows that the H land improvement district has made water management, which was
used turbidity monitoring system in the irrigation period of 2013. With the introduction of this system,
seems to reduce inflowing the concentration of radioactive substances to paddy field.
4. Conclusions
In this study, we introduced the turbidity monitoring system targeting the D irrigation system in
the Abukuma River basin. We also examined relationship between concentration of radioactive
substances and turbidity of agricultural water and the feasibility of reducing the inflow of radioactive
substances in paddy fields and canal system.
As a result, on the D irrigation system, relationship between concentration of radioactive
substances and turbidity of agricultural water is accumulated, the system of monitoring the turbidity of
agricultural water, and notify the administrator in real time has been shown the possibility of the
inflow prevention of radioactive substance.
The analysis revealed the following: 1) The concentration of radioactive substances in the
agricultural water is high when the turbidity of the agricultural water is high, there was a significant
correlation in the D irrigation system; 2) it is possible to deduce concentration of radioactive
substances in a simple manner by monitoring the turbidity; 3) continuous monitoring the turbidity is
required to reduce the radioactive substances flowing into the paddy area.
In order to introduce the turbidity monitoring system, such irrigation system inflow of
radioactive substances is expected to perform water management operation, verification of the
relationship between concentration of the radioactive substance and turbidity of agricultural water for
each region is required. Furthermore, by refraining from discharge to the main canal when it rains the
flow of turbid water is expected, the administrator of the irrigation system, can reduce the inflow of
radioactive substances.
As an issue in the future, further research is required to examine of inflow in irrigation system
through main canal. There is a need to analyze the case of more the turbidity monitoring system.
Acknowledgments
This study was carried out as part of the research agenda of the "Movement monitoring of
radioactive cesium in decontamination paddy" Ministry of Agriculture, Forestry and Fisheries
Research project commissioned. The authors express their gratitude to Ministry of Agriculture,
Forestry and Fisheries and the H land improvement district for their expensive cooperation.
References
Kubota T, Hitomi T, Hamada K, Yoshioka K, Sato M, Saito T, 2013, "Removal of radioactive Cs

from irrigation water at the inlet of a paddy field using chaff or other materials." Japan Technical
report of the national institute for rural engineering, 214, 123-133. (in Japanese with English
abstract)
Naka T, Wakasugi K, Haraguchi N, Okushima S, Shiono T, Ishida S, Yoshimoto S, Imaizumi M ,
2012, "Development of physical decontamination technologies farmland soil." Japan Water, land
and environmental engineering, 80(7), 19-22. (in Japanese)
Ministry of Agriculture, Forestry and Fisheries, 2013, Countermeasures about the current status of
radioactive material in reservoirs in Fukushima, Japan. Ministry of Agriculture, Forestry and
Fisheries, 1-20. (in Japanese)
Forestry and Forest Products Research Institute, 2012, Observations of radioactive substances in
stream water in August-October, http://www.ffpri.affrc.go.jp/press/2012/20121220/, September
2013. (in Japanese)
THE FUTURE
P-06
National Risk Assessment of Irrigation on the Farmland near

Wastewater Treatment Plants in KOREA
Jae-Ho Choi*, Chun Gyeong Yoon*, Han-Pil Rhee**, Moonsoo Cho*, Je ha Ryu*
*Department of Environmental Science, Konkuk University, Seoul, Korea
**ETwaters Inc., Seoul, Korea
ABSTRACT
Agricultural water supplies that are located nearby WWTPs (wastewater treatment plants) are more
than 130 places in Korea. The considerable part of the stream flow in these farmlands is dependent on
effluence of the facilities in dry season. It explains the indirect or direct reuse of effluence of WWTPs.
And, most of the farm workers use the effluent which does not have additional treatment because they
have lack of the knowledge about water reuse. In addition, there are short of review for health and
hygiene safety. Therefore, this study focused on the review of safety of those farmlands.
The 53 farmlands are located nearby WWTPs were investigated whether or not farm workers use
effluence of facilities as irrigation water on their paddy rice fields. Total coliform, fecal coliform,
E.coli and some heavy metals concentration of paddy water and soil were measured. The methods of
quantitative microbial and toxic risk assessment were applied for review of safety from wastewater
irrigation. This study used the E. coli concentrations in the microbial risk assessment for enteric
disease in the paddy fields. The acceptable level of annual microbial risk is less than 10-4. In toxic risk
assessment, acceptable carcinogenic risk ranges from 10-6 to 10-4, and non-carcinogenic risk can be
considered acceptable when lower than 1.
The result of the microbial risk analysis was 5.9×10-4. It doesn’t satisfy the minimum. Carcinogenic
risk was 4.01×10-5 and non-carcinogenic was 1.17×10-1. Those are not enough to get a safety,
because the values are close to the standard even though the study performed for only E.coli and some
toxic metals in the specific duration.
Keywords:
1. Introduction
Water has become an increasingly scarce resource worldwide. In 2025, two thirds of the world’s
population will be suffering moderate to high water stress, and half of the population will face real
constraints in their water supply (Lazarova et al., 2001). In South Korea, which has experienced
shortages in agricultural water supplies, wastewater reuse has been seriously considered as a way to
address such shortages. The Ministry of the Environment announced a plan to reuse 1200 million tons
of wastewater by 2016 (Ministry of Environment, Republic of Korea, 2007). If wastewater proves to
be a suitable alternative water source, problems of water shortage should improve. Wastewater reuse
for agriculture has already been applied in many countries around the world (U.S. EPA, 1992).
Wastewater reuse has the advantage of guaranteeing substitute water resources and increased
agricultural production, and can also help alleviate water pollution (Jung et al., 2005a; Jung et al.,
2005b). Wastewater treatment plant effluent is reused as irrigation water at more than 130 sites in
South Korea. In these areas, nearby wastewater treatment plants release water that is directly or
indirectly pumped into rice paddies. During the dry season, which is the most intensive agricultural
irrigation period in South Korea, a large portion of stream flow in such areas depends on discharge
volumes from wastewater treatment plants. Most farmers do not recognize that their irrigation water is
wastewater, while some farmers ignore the hazards by indiscreet reuse of wastewater.
However, wastewater treatment plant effluent usually contains various heavy metals and high
levels of microorganisms. Therefore, the quality and safety of wastewater must be guaranteed for both
farmers who may have direct contact with wastewater and the general public. Raw or partially treated
wastewater effluent could affect producers and consumers alike. Farmers are especially likely to have
direct contact with wastewater through ingestion and dermal contact (Peasey et al., 2000), and heavy
metals which are contained in farm products and soil could cause acute or chronic human health
problems. The U.S. EPA (1992) and the World Health Organization (WHO, 2000) have suggested
guidelines for the reuse of wastewater. And many quantitative risk assessments for wastewater reuse
have also been reported globally. However, while many studies exist on microbial risk assessment of
wastewater reuse, few findings are applicable to rice paddy irrigation with wastewater. Irrigation water
for rice paddies occupies the largest portion of agricultural water use in South Korea and paddy
products require large amounts of water. There is a significant chance of human contact with irrigation
water and consequently a high possibility of adverse human health effects if that water is
contaminated.
The objectives of this study were to evaluate the safety of irrigation water containing
microorganisms and toxic heavy metals. Microbial risk was estimated by analysis of Escherichia coli
(E. coli) concentrations to evaluate human health problems from enteric diseases. The quantitative
microbial risk assessment (QMRA) model was used to assume the infectious risk (Blumenthal et al.,
2000). Risk of toxic heavy metals was estimated by various exposure pathways and media, as well as
the risk value of each heavy metal present in irrigation water sources.
2. Material and Methods
2.1 Nationwide monitoring sites and sampling

This study selected 53 monitoring sites of farmland (specifically, paddy fields) located nearby
wastewater treatment plants from June to July 2009 (figure 1). The obtained irrigation water samples
were analysed for Escherichia coli (E. coli) and heavy metal concentrations. E. coli were determined
using standard methods (APHA, 2005). The density units were expressed as the most probable number
(MPN) per 100 mL. Heavy metals were detected by inductively coupled plasma-atomic emission
spectroscopy (ICP-AES). Target heavy metals from paddy soil and irrigation water were As, Pb, Ni,
Cu, Sb, Zn, Cd, and Cr.
Figure 1. 53 monitoring sites – farm lands nearby WWTP in South Korea
2.2 Quantitative Microbial risk assessment

This study used a method of quantitative microbial risk assessment that was previously
developed to assess water quality and set water quality standards for drinking water (Macler and Regil,
1993). The method consisted of four ordered steps: hazard identification, exposure assessment,
dose-response assessment, and risk characterization (Fewtrell and Bartram, 2001; Jung et al., 2005a).
Escherichia coli are the most common member of the fecal coliforms and a representative
microbiological indicator of water quality. The presence of E. coli also indicates the presence of other
pathogenic organisms (An et al., 2002). Hunter et al. (2011) used experimental data to derive
regression equations describing the relationships between the concentration of pathogens
(Cryptosporidiosis, giardiasis) and Escherichia coli Haas et al. (1999) introduced the QMRA method
using E. coli concentrations. Coliform bacteria concentration can be determined more easily than
assessing pathogenic organisms, and thus E. coli is often used as an indicator of water quality.
In this study, the exposure pathway for microorganisms was limited to ingestion by farm workers.
Exposure through inhalation and dermal contact was deemed less likely because rice paddies are
flooded during the irrigation period for an average of 100 days (Jung et al., 1999). Asano et al. (1992)
assumed that golfers are exposed to 1 mL of reclaimed water per day from handling golf balls and
staining of their clothes. Therefore, Jung et al. (2005a) proposed that farm workers ingest double the
amount of reclaimed water than golfers, because the opportunity for contact with pathogens in a rice
paddy field is greater than on golf courses and because of the aging farm-worker population in Korea,
which may mean that agricultural workers have weaker immune systems (Rhee et al., 2009; Jung et al.,
2005a). The Beta-Poisson dose-response model was used to quantify the risk of microbial ingestion, as
follows eq. 1(Haas et al., 1993).
I
N
PI = 1 − [1 + (2 α − 1)] (1)
N 50
Where PI is the risk of infection by ingesting pathogens in drinking water, N is the dose of
ingested microorganisms, N50 is the microbial dose resulting in 50% infection, and α is the slope
parameter. Human dose-response information was available when the exposure level was low, and the
best-fit dose-response parameters used for microbial ingestion were proposed to be N50 = 8.60 x 107
and α = 0.1778 (Haas et al., 1999). Eq. 2 represents the simplest assumption of the probability of
morbidity which considered the probability distribution, health, age, and working period of assessing
target
PD = PD :I × PI (2)
Where PD is the risk of an infected person becoming diseased, and PD:I is the probability of an
infected person developing clinical disease. To consider the probability of morbidity, the 50%
midpoint of the value was used in the calculation (PD:I = 0.5). The probability of farm workers
becoming infected was assumed to be two times greater than that of the control population (Jung et al.,
2005a) due to the age of agricultural workers and their high levels of infection possibility compared to
non-agricultural working adults (Nwachukul and Gerb, 2004).
The process of risk characterization combines information on exposure and dose-response into
an overall estimation of the likelihood of an adverse consequence (Burmaster and Anderson, 1994).
However, there are many sources of uncertainty. Therefore, a partial Monte Carlo simulation was
performed using best-fit point estimates of infectivity (N50, α ) and probability normal distributions of
E. coli concentrations. The Monte Carlo simulation also shows the relative contribution of variability
as a distribution (Tanaka et al., 1998).
2.3 Heavy metal toxic risk assessment

The quantifiable risk assessment used here was initially developed largely to assess human
health risks associated with exposure to chemicals (National Academy of Science, 1983). The process
is similar to that of microbial risk assessment and includes in order: hazard identification, exposure
assessment, dose-response assessment, and risk characterization (An et al., 2007; Fewtrell and Bartram,
2001; Jung et al., 2005a). South Korea’s Ministry of the Environment established risk assessment
guidance and published a Korean Exposure Factors Handbook (Jang et al., 2007; Ministry of the
Environment, Republic of Korea, 2006a, 2006b)
Farm workers are both producers and consumers. While at work, they can be exposed to
irrigation water and paddy soil through ingestion or dermal contact. The possibility of inhaling vapour
or dust containing target heavy metals is very low. Therefore, the inhalation pathway was excluded.
Receptor doses were calculated using a general equation. The toxic characteristics of target heavy
metals were based on the Integrated Risk Information System (IRIS) of the U.S. EPA (1996). Table 1
lists the proposed exposure factors for estimating receptor dose, and Table 2 details the exposure
mediums and pathways. Table 3 lists toxic characteristics of target heavy metals based on IRIS.
In the risk-characterization process, the total health risk to the population was estimated by
considering all pathways and analysing the contribution of each pathway. According to a proposal by
the U.S. EPA, allowable carcinogen risk ranges from 10-6 to 10-4 (U.S. EPA, 1991) and non-harmful
levels of non-carcinogenic risk are less than 1.
Table 1. Proposed exposure factors among South Korean
Agriculture (adults)
Parameter unit Reference
Carcinogen Noncarcinogen
1)
AT Average lifetime Days 10950 25550 KNSO, 2006
2)
ED Exposure duration years 30 70 NIER, 2006
U.S. EPA,
EF Exposure frequency days/year 100
1997
Jang et al.,
EFf Exposure frequency for food days/year 365
2007
2)
Bw Body weight Kg 60 NIER, 2006
SA Skin surface area cm2 17000 2)

NIER, 2006
U.S. EPA,
SAe Exposed skin surface area cm2 5700
2004a
U.S. EPA,
ABS Absorbed fraction from soil to skin unit-less As : 0.03, Cd : 0.001
2004a
U.S. EPA,
AF Soil to skin adherence factor mg/cm2 0.07
2004a
Jung et al.,
CRw Rate of irrigation water ingestion L/day 0.002
2005a
U.S. EPA,
CRs Rate of soil ingestion mg/day 100
1997
g/day-cap 1)
CRf Rate of food ingestion 215.9 KNSO, 2006
ita
FI Fraction of ingestion unit-less 1 -
Chemical-specific dermal U.S. EPA,
PC cm/hr 0.00084
permeability constant 2004a
ET Exposure time hr/day 8 -
Note) 1) KNSO: National Statistical Office of Korea, 2) NIER: National Institute of Environmental Research, Korea
Table 2. Exposure media and pathways in the paddy rice fields
Transport medium Exposure medium Exposure pathway

Irrigation water ingestion
Irrigation water (paddy water)
Dermal contact in water
Irrigation water Soil ingestion
Paddy soil
Dermal contact in soil
Air Inhalation
Table 3. Toxic characteristics of target heavy metals
RfD/RfC1) (mg/kg-day) SF2) (kg-day/mg) 5) 6)

BCFgrain PC
Oral Oralfood Oraldissolved Dermal Oral Dermal
As 3.0 × 10-4 - 2.9 × 10-4 1.5 × 100 6.1 × 101 0.036 1.0 × 10-3
Pb 3)
5.0 × 10-4 - - - 4)
8.5 × 10-3 - 0.012 1.0 × 10-4
Ni - - 2.0 × 10-2 8.0 × 10-4 - - 0.032 2.0 × 10-4
Zn 3.0 × 10-1 - - - - - 0.25 6.0 × 10-4
Cd 5.0 × 10-4 1.0 × 10-3 - 1.3 × 10-5 - - 0.364 1.0 × 10-3
Cr6+ 3.0 x 10-3 0.06 2.0 x 10-3
Note) 1) RfD: Reference Dose / RfC: Reference concentration, 2)SF: Slope factor, 3) Integrated Risk Information System, U.S.
EPA, 4) Korean National Institute of Environmental Research, 5) Chemical database of the California Environmental
Protection Agency, 6) U.S. EPA, 1996, Soil Screening Guidance: Technical Background Document, and 7) U.S. EPA, 2004,
Risk assessment guidance for superfund
3.1 Result of Quantitative microbial risk assessment

Irrigation water and paddy field of target sites were monitored from June to July 2009. The
coliform concentrations varied within each monitoring site. Table 4 shows coliform concentrations of
the study area where agricultural irrigation water was supplied from locations near wastewater
treatment plants. The Monte Carlo simulation was performed based on 10,000 trials, and risk values
with a 95% confidence interval were used.
Table 4. Coliform concentrations in the study area (MPN/100ml)
Paddy rice field Irrigation water

MEAN STDEV MEAN STDEV
Total Coliform 40271.7 130505.8 47471.0 97267.1
Fecal
22216.7 58489.2 21713.7 47960.3
Coliform
E. coli 3347.1 12453.9 8924.5 36110.73
The concentrations of each type of coliform in the paddy fields and irrigation water are shown
in figure 2, figure 3 presents results of the Monte Carlo simulation. Blue regions represent distribution
and frequency of positive risk values, and grey regions denote negative values.
Figure 2. Concentrations of each type of coliform in paddy fields and irrigation water
Figure 3. Microbial risk assessment by Monte-Carlo analysis
Table 5 shows the annual estimated risk value. Results of the microbial risk analysis
demonstrate that the average risk to farmers in the 53 monitoring sites is 5.9×10-4 in paddy fields and
1.6×10-3 in irrigation water. These annual risk values are insufficient because the values are close to
the standard limit, or higher than safety level. According to the microbial risk standard suggested by
the U.S. EPA, safety levels are less than 10-4. The range of 10-4 to 10-6 was considered a reasonable
level of risk for communicable disease transmission, and annual values above 10-4 were considered
high infection risks (Haas et al., 1993). Additionally, it should be noted that significant coliform
concentrations were observed during early irrigation periods (Rhee, 2009). Risk values of irrigation
sources in experimental plot scale paddy field were tabulated in Table 5 (21st Century Frontier
Research Program, 2011). In comparison with reported experimental data in plot scale, risk values of
paddy fields are between WWTP effluent and reclaimed water. National risk values cannot be
absolutely confirmed either safe or unsafe. Because it’s not result of census all about 130 farmlands
nearby WWTP and this study period was short. However, it is possible to grasp whether some level of
nationwide irrigation water. Thus, it is suggested that additional treatment should be considered for the
safe reuse of WWTP effluent on paddy rice field.
Table 5. Microbial risk values (Coliform)
Irrigation Reclaimed
Paddy field Ground water WWTP water
water water
Nationalwide monitoring sites Plot scale
Mean
Confidence interval
(95%)
3.2 Result of heavy metal toxic risk assessment

The concentrations of each heavy metal satisfy the South Korean water quality standards for
agriculture and FAO-recommended maximum concentrations. Tables 6 and 7 list levels of heavy
metals detected in irrigation water and paddy fields. Risk assessment for the heavy metal
concentrations measured in irrigation water and paddy soil was combined with exposure pathways.
This study period was in growing season for a comprehensive assessment. Farm product wasn’t
cropped in growing season. For this reason, the heavy metal concentrations of farm products was
calculated using the bioaccumulation factors (BCF) for the concentrations in rice paddy soil and the
rice factors noted in the Korean Exposure Factors Handbook (Ministry of Environment, Republic of
Korea, 2007), instead of directly measurement.
Table 6. Levels of heavy metals detected in paddy fields (mg/kg)
1) 1) Paddy rice field

Concern Level Action Level
MEAN STDEV.
As 6 15 N.D -
Pb 100 300 14.514 11.631
Ni 40 100 8.355 7.615
Cu 50 125 9.730 7.541
Sb - - 21.884 21.129
Zn 300 700 159.409 61.568
Cd 1.5 4 1.164 0.859
Cr - - 29.333 22.217
Note) 1) Standard level has suggested by Ministry of Environment, Republic of Korea.
Table 7. Levels of heavy metals detected in irrigation water (mg/L)
1) 2) Irrigation water
KSAW RMC
MEAN STDEV.
As 0.05 0.1 0.00738 0.00472
Pb 0.1 5 0.03111 0.14418
Ni - 0.2 0.00588 0.00569
Cu - 0.2 0.00650 0.01227
Sb - - 0.00661 0.00505
Zn - 2 0.08744 0.02218
Cd 0.01 0.01 0.00022 0.00015
Cr Cr6+: 0.05 0.1 0.01238 0.01340
1)
Note) KSAW: Korean Standard for Agricultural Water-quality (the Ministry for Food, Agriculture, Frestry and Fisheries
2
Republic of Korea), ) RMC: Recommended Maximum Concentration (FAO)
Table 8 lists the risk values associated with various exposure pathways. The total carcinogenic risk
value was 4.01 x 10-5, and the total non-carcinogenic risk value was 1.17 x 10-1. The total value of
carcinogen and non-carcinogen risk was lower than the U.S. EPA proposed value. According to a U.S.
EPA proposal, allowable carcinogenic risk levels range from 10-6 to 10-4, and allowable
non-carcinogenic risk levels are lower than 1 (U.S. EPA, 1991). Therefore, both carcinogenic risk and
non-carcinogenic risk associated with WWTP effluent were at acceptable levels. .
Table 8. Risk values associated with various exposure pathways
Exposure pathway Carcinogenic risk Non-carcinogenic risk

Water ingestion 4.13 × 10-8 4.05 × 10-4
Dermal contact in water 3.71 × 10-5 4.21 × 10-3
Soil ingestion 2.87 × 10-8 1.19 × 10-2
Dermal contact in soil - 1.33 × 10-2
Ingestion of rice 2.71 × 10-6 6.04 × 10-1
Total risk 3.99 × 10-5 6.34 × 10-1
3.3 Distribution of national risk values

The average risk values which are about E. coli and heavy metal are acceptable in safety level.
However, risk values have a regional variation. For this reason, we classified 3 categories by risk
value (Figure 4). Region Ⅰ is lower, Region Ⅱ is similar, and Region Ⅲ is higher than concern
level that was established by U.S. EPA. More than 90% of the study area is suitable for the safety level.
In microbial risk assessment, 2 sites were belonging to Region Ⅲ, and about 3.8 times higher than
safety level. In heavy metal risk assessment, 2 sites and 5 sites appertained in Region Ⅲ of
carcinogen and noncarcinogen risk, respectively. And that regional values about 1.6 times higher than
safety level. The caution sites are assessed that continuous management and additional treatment
reclamation of WWTP effluent for irrigation on farmland are needed. Also, in order to determine the
precise caution sites, more detailed and definitive studies are needed. Because, the monitoring duration
of this study was short and only one microorganism was examined
Figure 4. Distribution of national risk values by microbial and heavy metal
4. Conclusions
The safety of public health is an important issue in wastewater reuse. Quantitative risk
assessment offers an important tool for evaluating the risks associated with the use of reclaimed
wastewater. The objective of this study is quantitative risk assessment about E. coli and target heavy
metals in irrigation water originated by WWTP effluent in South Korea. The main purpose of this
study was to grasp the risk tendency for irrigation water that uncontrolled wastewater take a large
portion. Microbial risk was found to range from 10-3 ~ 10-4. These microbial risk levels suggest that
additional management and attention to the quality of irrigation water are needed. On the other hand,
carcinogenic mean risk levels of heavy metals through various exposure pathways were approximately
10-5 and non-carcinogenic mean risk levels were approximately 10-1. Both risk values were at
acceptable levels, but significant differences existed by region. In addition, the monitoring duration of
this study was short and only one microorganism was examined. Thus, more detailed and definitive
studies are needed to collect data on various pathogenic microorganisms and toxic chemicals.
Acknowledgement
This work was supported by the ‘Ag-BMPs development project for water quality improvement
in Saemangeum estuarine reservoir’ funded by the Ministry for Agriculture, Food and Rural Affairs,
Republic of Korea (MAFRA).
References
American Public Health Association (APHA), 2005, In: Greenberg, A.E., Trussell, R.R., Clisceri, L.S.
(Eds.). Standard methods for examination of water and wastewater. 21th Ed., Washington, DC.,
USA.
An, Y. J., Kampbell, D. H., and Breidenabch, G. P., 2002, Escherichia coli and total coliforms in
water and sediments at like marinas. Environmental Pollution, 120, 771-778
An, Y. J.., Yoon, C.G., Jung, K.W., Ham, J.H., 2007, Estimating the microbial risk of E. coli in
reclaimed wastewater irrigation on paddy field. Environ. Monit. Assess. 129, 53-60
Asano, T., Leong, L. Y. C., Rigby, M. G. and Sakaji, R. H., 1992, Evaluation of the California
wastewater reclamation criteria using enteric virus monitoring data. Water Science and
Technology. 26(7-8), 1513-1524.
Blumenthal, U. J., Mara, D. D., Peasey, A., Ruiz-Pakacios, G., and Stott, R., 2000, Guidelines for the
microbiological quality of treated wastewater used in agriculture: recommendations for revising
WHO guidelines. World Health Organization, Special Theme-Environment and Health. 78,
1104-1116
Burmaster, D. E. and Anderson, P. D., 1994, Principles of good practice for use of Monte Carlo
techniques in human health and ecological risk assessment. Risk Analysis, 14(4), 477-481.
Committee on the Institutional Means for Assessment of Risks to Public Health, National Research
Council., 1983, Risk assessment in the federal government: managing the process. National
Academy of Sciences, Washington, DC.
Fewtrell, L. and Bartram, J., 2001, Water quality. Guidelines, standards and health: assessment of risk
and risk management for water-related infectious disease. IWA Publishing, World Health
Organization (WHO).
Haas, C. N., Rose, J. B., and Gerba, C., 1999, Quantitative microbial risk assessment. Wiley, New
York.
Haas, C. N., Rose, J. B., Gerba, C. and Regli, S., 1993, Risk assessment of virus in drinking water.
Risk Analysis, 13, 545-552.
Hunter, P. R.,Sylor, D., Andrele, M., Risebro, H. L., Nichols, G. L., Kay, D. and Hartemann, P., 2011,
Quantitative Microbial Risk Assessment of Cryptosporidiosis and Giardiasis from Very Small
Private Water Supplies. Risk Analysis, 31, 228-236.
Jang. J.Y., Jo, S.N., K, S.Y., and Cheong, H.K., 2007, Korean exposure factors handbook. Ministry of
Environment Republic of Korea. Seoul, Korea. (in Korean)
Jung, H. W., Kim, S. J., Kim, J. S., An, B. K., Lee, K. H., Lee, N. H. and Jung, S. O., 1999, Irrigation
and Drainage Engineering. Dongmyung-sa, Seoul, Korea. (in Korean)
Jung, K. W., Yoon, C. G., An, Y. J., Jang, J. H., and Jeon, J. H., 2005a, Microbial risk assessment in
treated wastewater irrigation on paddy rice plot. Korea J. Limnol., 38(2), 225-236. (in Korean)
Jung, K. W., Yoon, C. G., Jang, J. H., Kim, H. C., and Jeon, J. H., 2005b, Investigation of indicator
microorganism concentrations after reclaimed water irrigation in paddy rice plots. Journal of the
Korean Society of Agricultural Engineers, 47(4), 75-85. (in Korean)
Lazarova, V., Levine, B., Shack, J., Cirelli, G., Jeffrey, P., Muntau, H., Salgot, M. and Brissaud, F. ,
2001, Role of water reuse for enhancing integrated water management in Europe and
Mediterranean countries. Water Science and Technology, 43(10): 25-33.
Macler, B. A. and Regil, S., 1993, Use of microbial risk assessment in setting US drinking water
standards. Int. J. Food Microbiol, 18:245-256.
Ministry of Environment Republic of Korea., 2006a, Guidance about risk assessment for soil
contamination. (in Korean)
Ministry of Environment Republic of Korea, National Institute of Environmental Research., 2006b,
Notification about risk assessment method. (in Korean)
National Academy of Science., 1983, Risk Assessment in Federal Government: Managing the Process.
National Academy Press, Washington, DC.
Nwachukul, N. and Gerb, C. P., 2004, Microbial risk assessment: don't forget the children. Curr. Opin
Microb., 7:206-209.
Peasey, A., Blumenthal, U. J., Mara. D., and Ruiz-Palacios, G., 2000, A review of policy and
standards for wastewater reuse in agriculture: a Latin American perspective. WELL Study,
London School of Hygiene & Tropical Medicine, WEDC Loughborough University, UK., 68(2),
15-18.
Rhee, H. P., Yoon, C.G., Jung, K. W., and Son, J. W., 2009, Microbial risk assessment using E. coli in
UV disinfected wastewater irrigation on paddy. Korea Society of Environmental Engineers,
14(2), 120-125. (in Korean)
Tanaka, H., Asano, T., Schroeder, E. D. and Tchobanoglous, G., 1998, Estimating the safety of
wastewater reclamation and reuse using enteric virus monitoring data. Water Environment
Research, 70(1):39-51.
U.S. EPA., 1989, Risk assessment guidance for superfund volume I: human health evaluation manual
(Part A). EPA No. 540/1-89/002. Office of Emergency and Remedial Response U.S.
Environmental Protection Agency, Washington, D.C. 20450
(Part B, development of risk-based preliminary remediation goals). EPA No. 540/R-92/003.
Office of Emergency and Remedial Response U.S. Environmental Protection Agency,
Washington, D.C. 20460
U.S. EPA., 1992, Guidelines for water reuse, EPA No. 625/R-92/004. US Agency for International
Development. 9 Aug. 2004.
U.S. EPA., 1996, Soil Screening Guidance: Technical Background Document. EPA No. 540/R-95/128.
Office of solid Waste and Emergency Response. U.S. Environmental Protection Agency,
Washington, DC.
(Part E, supplemental guidance for dermal risk assessment). EPA/540/R/99/005. Office of
Superfund Remediation and Technology Innovation U.S. Environmental Protection Agency
Washington, D.C.
WHO., 2000, Guidelines for the microbiological quality of treated wastewater used in agriculture
recommendations for revising WHO guidelines: Special Theme-Environment and Health. World
Health Organization. Geneva, Switzerland.
21st Century Frontier Research Program., 2011, Application of Integrated Technologies for
wastewater reclamation and reuse system for agriculture. (in Korean)

PAWEES 2013 Full Paper

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PAWEES 2013 Full Paper

Uploaded by

Copyright:

Available Formats

Contents

A-06 Climate Change Action Plan for Water Resources in Taiwan

D-08 Investigation of Organic Fertilizer to Reduce Insecticide - Assessment of Paddy Ecosystem

F-08 Analysis of Irrigation Service Needs by Rice Farming Families in Japan

F-10 Runoff Characteristics of Non-point Source Pollution from Reclaimed Paddy

P-04 Development of Irrigation Management Method for Reducing Inflow of Radioactive

Impact of Average Elevation of Basins on Earlier Snowmelt

ITO Nobuo, SUTO Yuji, NAKAMURA Kazumasa

Snowmelt runoff starts Warming accelerates decrease in

Demand for paddy-field

Mar. Apr. May Jun. Jul. Aug. Sep.

500～600 Headworks C (area weighted

600 Increase in precipitation per 100m elevation (warm period)

300 Ratio of the two

Figure 5 Method of correcting precipitation values

Table 1 The 9 climate models used for analysis

Black bold line：Average of present situation (1980-1999)

Figure 6 Results of runoff analysis

3. Results and consideration

Value of present state obtained

3/27 4/1 4/6 4/11 4/16 4/21 4/26

Figure 7 Change in peak runoff day

Value of present state obtained

Figure 8 Change in cumulative volume of runoff in the snowmelt season (February-May)

Predicted value obtained

Figure 9 Change in the total volume of runoff in irrigation period (May-August)

Assessment of Climate Change Impact on Crop Yields in

RAY-SHYAN WU i*, MING-HSU LI**, JI-TANG FANG***, CHI-MEI WANG****

1.2 Literature review

GDD = ∑((Tmax + Tmin )/2 − b) (1)

2.2 Weather generator model

P = P�ȷ × (− ln(1 − N) (3)

2.3 Irrigation demand estimation

ETi = ET0i × K c (4)

2.4 The theory of Principal Component Analysis (PCA)

x11 x12 … x1m

3. Climate change scenario

Temperature (°C) Rainfall (%)

CTm = Tm + ∆Tm (7)

CPm = Pm × RPm (8)

4. Result and discussion

4.1 Crop yield simulation using the DSSAT model

Table 2. The test and verification of DSSAT model

Simulation yield (DSSAT)

2007 5288 5441 3.77%

2008 5416 5620 2.89%

4.2 Principal Component Analysis

Model Test Value

Table 4. Eigenvalue and explanation of variance of models

Principle Explanation of variance Accumulation of explanation

Table 5. The component loading matrix of models

SRAD_B -0.45 0.33 -0.54 ---

SRAD_B 0.81 -0.26 0.30 0.41

SRAD_B 0.87 0.08 0.11 0.46

SRAD_B 0.84 -0.01 0.26 0.45

Baseline TEMP_A TEMP_B SRAD ---

Climate Change Action Plan for Water Resources in Taiwan

Wei-Fuu Yang*, Chi-Ming Chen**, Pei-Jung Wu***

To respond to climate change impact, to strengthen adaptive capacity, to lower social

Table 1 Adaptation strategies and corresponding measures in water resources sector

Adaptation Strategies Adaptation Actions

Overseeing all developmental or construction projects. While

2. Water resources in Taiwan

RAY-SHYAN WU i, MING-HSU LI, JI-TANG FANG, CHI-MEI WANG****

Wei-Fuu Yang, Chi-Ming Chen, Pei-Jung Wu

Hajime Tanji, Katsuhiro Sakurai, Ataru Nakamura, Hirohide Kiri*, Tetsuo