pawees 2013 full paper

[1]

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

Contents

[2]

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

[3]

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

Session 1

PAWEES 2013 (12TH) INTERNATIONAL CONFERENCE ON AGRICULTURAL WATER AND RURAL ENVIRONMENT FOR THE FUTURE 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

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

A-02

2 PAWEES 2013 (12TH) INTERNATIONAL CONFERENCE ON AGRICULTURAL WATER AND RURAL ENVIRONMENT FOR THE FUTURE

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.

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.

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

Storage volume

River discharge

Demand for paddy-field irrigation water

Reservoir volume

Snowmelt runoff starts earlier due to warming.

Warming accelerates decrease in the amount of stored water.

Storage in a reservoir might not be sufficient due to warming.

Figure 1 Conceptual diagram of how warming affects paddy field irrigation in Hokkaido


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.

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

Figure 4 Straight 4-stage storage tank model b4

H1

H2

H3

H4a5(z5)

b1

b2

b3

a1(z1)

a2(z2)

a3(z3)

a4(z4)

Figure 2 Inlet works and the size of the basin

Basin of Dam A(62.6km2)

Basin of Dam B(12.4km2)

Basin of Headworks C(80.7km2)

Dam A (area weighted average elevation : 280m）

Headworks C (area weighted average elevation : 261m ）

Dam B (area weighted average elevation : 222m）

0～100

700～800

600～700

500～600

400～500

300～400

100～200

200～300

0 10 20 30 40 50 60Proportion of area(%)

Elev

atio

n(m

)Figure 3 Elevation distribution of the basin area

containing the studied facility points


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.

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

Temperature ：-0.6℃/100m(elevation)

Corrected precipitation value＝daily precipitation observed at the nearest AMeDAS observation station ×1.6×1.5

0

100

200

300

400

500

600

700

0 100 200 300 400 500 600 700

Cumulative precipitation observed at the nearest AMeDAS observation station （mm）

Ratio of the two

Increase in precipitation per 100m elevation (cold period)Coefficient ＝

Increase in precipitation per 100m elevation (warm period)

Sum

of i

nflo

w in

to th

e da

m a

nd th

e ev

apot

rans

pira

tion

in th

e ba

sin（

mm）

Table 1 The 9 climate models used for analysis

Figure 5 Method of correcting precipitation values

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)


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.

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

Figure 6 Results of runoff analysis

Headworks CDam A

Dam B

Black bold line：Average of present situation (1980-1999)Other lines：Future forecast of each climate model




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.

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.

Dam A

Headworks C

Dam B

Average of predicted value and standard deviation

Predicted value obtained from the model

Value of present state obtained from AMeDAS data

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

Figure 7 Change in peak runoff day


Predicted value obtained from the model

Value of present state obtained from AMeDAS data

Dam A

Headworks C

Dam B

%80 90 100 110

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


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 from the model Value of present state obtained

from AMeDAS data

Dam A

Headworks C

Dam B

%50 60 70 80 90 100 110

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

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


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

Assessment of Climate Change Impact on Crop Yields in

Northern Taiwan by Using Principal Component Analysis

RAY-SHYAN WUi

*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,

*, MING-HSU LI**, JI-TANG FANG***, CHI-MEI WANG****

[email protected]. **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

A-03

mailto:[email protected]�


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 using the first order Markov chain (eq. 2). Ti was the ith day’s temperature; T�j was the jth month’s

average temperature; ρj was the first order serial correlation coefficient of Ti and Ti−1 in the jth

month; Ni was a random number between 0 and 1; and σj was the standard deviation of historical

data in the jth month.

Ti = T�j + ρj�Ti−1 − T�j� + Niσj

�1 − ρj

2 (2)

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), Pȷ� was rainy day average rainfall at the jth month (cm), and N was a random number between 0 and 1.

P = Pȷ� × (− 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 × Kc (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 nm original data matrix P (eq. 5).

P = �

x11 x12x21 x22

… x1m… x2m

⋮ ⋮xn1 xn2

⋱ ⋮… xnm

� (5)

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 (λi) and eigenvectors of the matrix. The principal component was the new

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).

ℒij =𝒲ij�λi𝒮j

= ryixj (6)

The variable ℒij was the jth variable on the ith component loading; 𝒲ij was the jth variable

on the ith component weight; λi was the ith eigenvalue of the component; 𝒮j was the standard

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 (%)

month CNRM_CM3 CSIRO_MK3_0 Ensemble CNRM_CM3 CSIRO_MK3_

0 Ensemble

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

year Real yield 1) (kg/ha) Simulation yield (DSSAT) (kg/ha) Deviation (%)

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).

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).

-6

-5

-4

-3

-2

-1

0 系集 CM3 MK3_0

Yie

ld C

hang

e(%

)

Ensemble


Table 3. KMO and Bartlett test

Model Test Value

Baseline Kaiser-Meyer-Olkin MSA 0.506 Bartlett Sphericity Test Significance 0.000

Ensemble Kaiser-Meyer-Olkin MSA 0.535 Bartlett Sphericity Test Significance 0.000

CM_3 Kaiser-Meyer-Olkin MSA 0.508 Bartlett Sphericity Test Significance 0.000

MK3_0 Kaiser-Meyer-Olkin MSA 0.562 Bartlett Sphericity Test Significance 0.000

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

Model Principle component Eigenvalue Explanation of variance

(%) Accumulation of explanation

variance (%)

Baseline

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

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

Ensemble

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

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

CM3

1 3.23 40.42 40.42

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

MK3_0

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

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

Crop yield variable Component

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


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


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


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

Model Component

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

http://tccip.ncdr.nat.gov.tw/NCDR/main/index.aspx�





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

A-06


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

1

Taking sustainability as the highest guiding principle for the nation's water resources management, and also focusing on protecting the aqua environment.

1-1

Overseeing all developmental or construction projects. While analyzing the costs and benefits and evaluating the influence on the environment, it is impor tant to also consider climate change and the nation's water resources to prevent the current aqua environment from being upset.

1-2

Systematically planning and managing the governance of river basins. The planning and management should be based on each river basin's special characteristics, while taking the aqua environment pollution control, sustainability of freshwater resources, and protection of biodiversity and ecosystems into consideration.

2

Reviewing water resources management from the supply side and reinforcing the efficient usage of water resources.

2-1 Revitalizing current water storage capacities. Repairing and maintaining related facilities when needed, mostly to prevent unnecessary loss of water through leakage or during transportation.

2-2

Implementing the reservoir watershed land use management, proper use of water resource operating funds to promote reservoir watershed conservation. Converting farmland into wo o dlands to prevent agricultural pollutants like fertilizers and pesticides from contaminating reservoir waters.

2-3 Reinforcing cross-regional conjunction use of surface and ground 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.

3 Establishing a database of the regional total quantity of water supply

3-1 Adjusting the water price to a reasonable level, changing water use patterns, establishing a reasonable, fair, and flexible transformation mechanism for water utilization, promoting water conservation efforts,


system and reviewing water resource total quantity control policies from the demand side in order to reinforce the efficiency of water resources.

reexamining the current architecture laws, strengthening the regulation of public buildings and facilities, and promoting the installation of water-conserving devices in public properties.

3-2 Encouraging the development of low waterconsumption industries. Products that require high water-consumption during production can potentially be imported from water-abundant nations.

3-3

Adjusting the agricultural system by considering the environmental sustainability and farmland productivity. If the two conditions are sustained, we can then promote precise irrigation and improve irrigation methods, which would improve rainwater resource efficiency and reduce the demand for irrigation.

4

Promoting water sustainability in line with the United Nations' water footprint concept.

4-1 Requiring products to be printed with water consumption labels for the consumers' reference to reduce the consumption of products with high water usage.

4-2 Encouraging corporations to establish waterconserving processes of production in order to decrease the water consumption.

4-3 Creating financial incentives for conserving and recycling water resources.

4-4

Calculating water accounts through systematic analysis using the concept of material flow and water balance. Examining the reasonableness of the monitoring data of river basins from dif ferent 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.

Sea

Water Resources Adaptation

Surface water

Water storage in reservoirs

Groundwater

Groundwater abstraction

Assessing water in reservoirs

Industrial consumption

Water purification

Water transportation

Agricultural consumption

Domestic consumption

Development and conservation of water resources

Effective water supply

Water demand

Import and export

PrecipitationEvaptranspiration

IssueI

IssueII

IssueIII

Assessing water in rivers

IssueIV

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 basinRainfall (mm)

Rainfall (mm) Baseline

Rainfall (mm)Tsengwen river basin

Baseline

Kaopingriver basin

BaselineA1B

A1B

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

3 key issues and 11 sub-issues

4 climate change factors

Rising temperature

Decreasing total rain days

Increasing rainfall

intensity

Rising sea level

Development and

conservation of water

resources

Surface water amount of surface water storage

amount of surface water storage

- -

Groundwater - amount of

groundwater storage

- amount of groundwater storage


Water storage in reservoirs

amount of water storage in reservoirs

amount of water storage in reservoirs

water turbidity and sedimentation

-

Water accessibility of reservoirs

- - water turbidity -

Water accessibility

of rivers

- amount of stream flow

water turbidity -

Water purification

water quality water quality water turbidity -

Water demand


amount of agricultural water demand

amount of irrigation water demand

- -


amount of domestic water demand

- - -


amount of industrial water demand

- - -

Decreasing total rain days

Increasing rainfall intensity

Surface water

Ground-water

Water accessibility of reservoirs

Water accessibility

of rivers

Assessing water in

rivers

Water purification




Increasing sedimentation

Increasingturbidity

Rising sea level

Intrusion of saltwater

Rising temperature

1 2 3

4

5

6

7 8

9

10

12

13

14 15 1611

1 2 3 4 5 6 7 8 9

Climate change factors

16 main impacts of climate change9 main sub-issues of water resources

Water quality deterioration

Decreasing stream flow

Increasing evapotranspiration

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.

L

M

H

H M LWater resources adaptation

Vulnerability

High possibility of significant

hazards

Possible hazardsneed follow

-up studies

H

H

M

H

M

L

M

L

L

Risk= hazard x vulnerability Hazard: the possibility of

significant impacts

Vulnerability: the consequences

of impacts

Climate change factors

Hazard

Risk

Low possibility

of significant hazards

Influenced extent and scale increase, no adaptive measures are implemented

Influenced extent and scale increase, adaptive measures have been implemented

Influenced extent and scale do not increase

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.

WRA, EPA, COA, MOEA

WRA, MOEA, NSC

Taking sustainability as the highest guiding principle for the nation's water resources management, and also focusing on protecting the aqua environment.

Development and conservation of water resources

Reviewing water resources management from the supply side and reinforcing the efficient usage of water resources.


Establishing a database of the regional total quantity of water supply system and reviewing water resource total quantity control policies from the demand side in order to reinforce the efficiency of water resources.

Water demand

Promoting water sustainability in line with the United Nations' water footprint concept.

Import and export of water sources

Action Plans for Water Resources Adaptation

National Climate Change Adaptation Policy

WRA, MOI, COA, MOEA

WRA, EPA, COA, MOEA, NSC

CEPDResponsible authorities

WRA

Strategy 1

Responsible authorities

Strategy 2

Strategy 3

Strategy 4

Related authorities

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

Classification NO. Action plans Responsible authorities

Decreasing surface water resource

1 Enhancement of mechanism and strategies of water resources under climate chaege WRA

2 Improvement and management of water environment and master plan EPA

Decreasing reservoir storage

3 Reinforcement and improvement projects for water-storage constructions (phase II)) WRA

4 Remediation project for reservior and catchment of Shimen WRA

5 Demostration project for low impact development planning in catchment of reservior WRA

Increasing of water demand for agriculture

6 Improvement of management of water right WRA

7 Planning the promotion policy for efficiency of uses in argricultural water resources COA

8 Saving agricultural consumption by promoting dryland irrigation piping and modernizing water-saving facilities COA

9 Promoting water-saving irrigation and adjusting farming mechanism in subsidence area of Changhua and Yunlin counties COA

Important integrated plans

10 Strategy planning and promotion of water fee for sustainable operation of drinking water industry WRA

11 Compilation of the storm water storage standard and legislation of retention facilities of buildings MOI

12 Planning of regulation pools during rezoning agricultural land COA

13 Establishment of legislation and reward mechanismof the water re-use WRA

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


PAWEES 2013 (12TH) INTERNATIONAL

CONFERENCE ON AGRICULTURAL WATER AND RURAL ENVIRONMENT FOR

THE FUTURE RAMADA PLAZA HOTEL, Cheongju, KOREA


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

B-04


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.

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

Country Irrigated area (ha)

Area affected by salinization (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.

Figure 1. Location of Uzbekistan

Uzbekistan

Google Map


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.

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

Figure 2. Location of Syrdarya region

Syrdarya region

Tashkent

Google Map

Figure 3. Average monthly temperature and rainfall in the Syrdarya region (2004–2012)

0.0

10.0

20.0

30.0

40.0

50.0

60.0

0.0

5.0

10.0

15.0

20.0

25.0

30.0

1 2 3 4 5 6 7 8 9 10 11 12

Rai

nfal

l (m

m)

Tem

pera

ture

(℃)

(Mon)

Rain (mm) Temp (℃)


ADF method

Normal method

Figure 4. ADF method

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.

(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.

50 m 100 m 0 m

Root zone

50 m 100 m 0 m

Root zone 1st infiltration loss 2nd infiltration loss

1st water supply 2nd water supply

Normal irrigation

Large infiltration loss

Normal method

Surge flow method

Figure 5. Simplified surge method


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.

pLC C

DMfTRAM 1)( ××−= (1),

Figure 6. Design of the experimental plot

Inlet

Weir

100m

Con

trol

AD

FSu

rge

Com

bina

tion

(AD

F +

Surg

e)

Con

vent

iona

l

Con

trol

AD

F

Surg

e

Com

bina

tion

(AD

F +

Surg

e)

Con

vent

iona

l

Earth canal

Con

trol

AD

FSu

rge

Com

bina

tion

(AD

F +

Surg

e)

Con

vent

iona

l

Concrete canal

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

0.9 m

Plot 7.2 mNext plot Border 1.8 m

8 FurrowsIrrigation according to treatment

2 FurrowsNo irrigation


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.

ECwECeECwLR

−=

)(5 (2)

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).

LRTRAMAW−

=1

(3),

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).

f

if W

WAWID ×= (4),

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)).

nif ctD = (5),

where Df is the infiltration water depth (mm); c and n are constants of the Kostiakov equation; and

it 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).

11

βαα Lta = (6),

where at is the elapsed time of water advance (s); 1α and 1β are constants of the water advance equation; and La is the distance water advanced from the inlet in the furrow (m).

22 βα += rr Lt (7), where rt is the elapsed time of water recession (s); 2α and 2β 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).

nar

ftt

cD

−

=60

(8),

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.

10002

1

×

+

=

+ WDD

S

fjfj

j (9),

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.

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.

Figure 8. Average irrigation time in one furrow

0

20

40

60

80

Cont Conv ADF Surge A+S

No.1 No.2 No.3 No.4 No.5

(min) Trial 1



Trial 2


No.13 No.14 No.15 No.11 No.12

Trial 3

Figure 7. Average amount of irrigation water in one furrow



Trial 2


No.13 No.14 No.15 No.11 No.12

Trial 3

0

2

4

6

8

10



(m3) Trial 1


Trial 1 Trial 2 Trial 3

Figure 9. Cumulative infiltration and elapsed time

y = 2.4354x0.4221

R² = 0.9444

0

10

20

30

40

0 40 80 120 160 200 240

(mm) No.1 Control

y = 1.9005x0.509

R² = 0.9596

0 40 80 120 160 200 240

No.10 Control

y = 1.8339x0.5016

R² = 0.9556

0 40 80 120 160 200 240

No.13 Control

y = 1.6376x0.5864

R² = 0.9975

0

10

20

30

40

0 40 80 120 160 200 240

No.2 Conventional

y = 1.7829x0.4581

R² = 0.9669

0 40 80 120 160 200 240

No.6 Conventionaly = 2.0495x0.5095

R² = 0.9745

0 40 80 120 160 200 240

No.14 Conventional

y = 1.8019x0.4687

R² = 0.9942

0

10

20

30

40

0 40 80 120 160 200 240

No.3 ADF

y = 1.7499x0.4101

R² = 0.9658

0 40 80 120 160 200 240

No.7 ADFy = 1.9012x0.5242

R² = 0.9622

0 40 80 120 160 200 240

No.15 ADF

y = 1.7049x0.4955

R² = 0.9627

y = 1.3263x0.5035

R² = 0.97650

10

20

30

40

0 40 80 120 160 200 240

No.4 Surgey = 3.368x0.4403

R² = 0.9793

y = 0.8977x0.6138

R² = 0.9448

0 40 80 120 160 200 240

No.8 Surgey = 2.3905x0.4805

R² = 0.9661

y = 1.5692x0.4572

R² = 0.9721

0 40 80 120 160 200 240

No.11 Surge

y = 1.1255x0.4323

R² = 0.9646

y = 1.0158x0.5347

R² = 0.9662

0

10

20

30

40

0 40 80 120 160 200 240

No.5 ADF+Surge

y = 1.7698x0.4735

R² = 0.9592

y = 1.1647x0.451

R² = 0.9851

0 40 80 120 160 200 240

No.9 ADF+Surge

y = 1.3643x0.4634

R² = 0.9619

y = 0.7655x0.5111

R² = 0.9872

0 40 80 120 160 200 240(min)

No.12 ADF+Surge

Before irrigation After irrigation


Trial 1

Water advance Water Recession

Trial 2 Trial 3

Figure 10. Water advance and recession and elapsed time

y = 7.0639 x1.1543

R² = 0.9954

y = 22.327x + 6950.2R² = 0.2572

0 50 100

No.10 Control

y = 11.7296 x1.0782

R² = 0.9741

y = 72.618x + 10089R² = 0.6792

0 50 100

No.13 Control

y = 10.871x1.1417

R² = 0.9268

y = 59.844x + 7447.6R² = 0.7504

0

5000

10000

15000

0 50 100

No.2 Conventional

y = 9.0723x1.1547

R² = 0.968

y = 42.895x + 6133.4R² = 0.7856

0

5000

10000

15000

0 50 100

No.3 ADF

y = 16.909x1.0394

R² = 0.9748

y = 16.274x + 4375.5R² = 0.043

0

5000

10000

15000

0 50 100

No.4 Surge 50m

y = 2.6946x1.3689

R² = 0.9857

y = 22.765x + 4821.6R² = 0.2421

0

5000

10000

15000

0 50 100

No.4 Surge 100m

y = 13.52x0.931

R² = 0.9948

y = -13.314x + 3755.5R² = 0.0602

0

5000

10000

15000

0 50 100

No.5 ADF+Surge 50m

y = 6.0291x1.1259

R² = 0.9686

y = 10.727x + 7169R² = 0.0434

0

5000

10000

15000

0 50 100

No.5 ADF+Surge 100m

y = 14.744x1.0473

R² = 0.9975

y = 48.307x + 4539.2R² = 0.9199

0 50 100

No.6 Conventional

y = 7.4989x1.1249

R² = 0.9902

y = 48.199x + 3838.1R² = 0.7605

0 50 100

No.7 ADF

y = 13.437x1.0824

R² = 0.9896

y = 14.908x + 2340.9R² = 0.7373

0 50 100

No.8 Surge 50m

y = 6.5853x1.082

R² = 0.9823

y = 40.294x + 5421.8R² = 0.8789

0 50 100

No.8 Surge 100m

y = 14.866x0.892

R² = 0.9772

y = 13.261x + 2050R² = 0.1939

0 50 100

No.9 ADF+Surge 50m

y = 4.2163x1.1144

R² = 0.99

y = 83.65x + 2838.3R² = 0.8316

0 50 100

No.9 ADF+Surge 100m

y = 4.595x1.2675

R² = 0.985

y = 42.796x + 3644.4R² = 0.9065

0 50 100

No.14 Conventional

y = 8.3263x1.2375

R² = 0.9806

y = 24.588x + 5267.5R² = 0.7116

0 50 100

No.15 ADF

y = 14.298x1.004

R² = 0.9907

y = 17.254x + 4025.1R² = 0.0521

0 50 100

No.11 Surge 50m

y = 4.4387x1.2195

R² = 0.9909

y = 45.252x + 6770.4R² = 0.8495

0 50 100

No.11 Surge 100m

y = 14.122x0.991

R² = 0.9994

y = 31.362x + 2906R² = 0.8894

0 50 100

No.12 ADF+Surge 50m

y = 6.1802x1.0829

R² = 0.9877

y = 5.3064x + 5496.6R² = 0.1054

0 50 100 (m)

No.12 ADF+Surge 100m

y = 12.7219 x1.0515

R² = 0.9876

y = 58.798x + 7560.5R² = 0.6945

0

5000

10000

15000

0 50 100

(s) No.1 Control


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.

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

3.57 3.57 3.57 3.57 3.57

0.18 0.12 0.01 0.07 0.03

No.13 No.14 No.15 No.11 No.12


Trial 3

Figure 11. Amount of water for necessary infiltration and subsequent infiltration loss

Necessary infiltration (39.6 mm) Infiltration loss

3.57 3.57 3.57 3.57 3.57

0.03 0.05 0.02 0.22 0.74



Trial 2

3.57 3.57 3.57 3.57 3.57

0.07 0.29 0.05 0.31 2.26

0.00

2.00

4.00

6.00



Infil

trat

ion

wat

er (m

3 )

Trial 1


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.

http://www.fao.org/docrep/003/t0234e/T0234E03.htm#ch2.4.2�

http://www.fao.org/docrep/s8684e/s8684e04.htm#chapter%203.%20furrow%20irrigation�





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

B-07


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.

Kiso River

Main canal

Kanayama HWInuyama HW

Supplied area

Irrigated land

PondsHW(Headworks)

Nagoya city

Makio Dam

5km

Ikeda branch area

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

Total sold price

10a

30a

5a

150,000 500,000 yen

Sellingfarmer

No farmerlandowner

Out of survey

Self-sufficientfarmer

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.

0

5

10

15

20

25

30

35

40

45

Narsring Paddling Tranplanting Pest control Clutivation Drying

Whole Japan Gifu Aichi Mie%

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

Year Law LandOwner

TotalArea

TotalValue

Cultivator Entity

1899 Arable land readjustment >2/3 >2/3 >2/3 Individuals, Group1909 Arable land readjustment >1/2 >2/3 >2/3 Individuals, Group, Organization1949 Land Improvement >2/3 Organization (WUA)

Table 2 a and b Laws related to farmland adjustment and irrigation and drainage Before the warProject Irrigation and drainage project Areal land readjustment projectLaw Irrigation association law Areal land readjustment lawAgreement none >2/3 of area, value, landownersCompulsory power none tax collectionOrganization Irrigation association land readjustment association

After the warProject Irrigation and drainage project Areal land readjustment projectLawAgreementCompulsory powerOrganization

land improvement law>2/3 of cultivators

limit propertyWater 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

Cultivator Cultivator Cultivator

Land owner Land owner Land owner


Land owner Land owner Land owner


Land manager

Before the war

After the war

Recently

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

320

40035090

84065

74

54

55

70

Others

Figure5 Breakdown of cultivated land in the Ikeda WMU (unit a)

(Numbers in the legend are ages of farmers.)

After the war

Land owner

Land owner

No cultivator

Land manager

Land owner Cultivator


CultivatorPresent

Land owner

Land owner

No cultivator

Land manager


CultivatorFuture

210 farmers

100 farmers

6 farmers

100 farmers

10 farmers

3 farmers

200 farmers

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.

0 20 40 60 80 100

020

4060

8010

0

CumPerPercent

Cum

TAre

a2P

er

Cumulative person

20

4

0

60

Cumulative area

ZZZ

YY

XC2

XC1XB1

XB2

XA１

XA2

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





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,

C-01


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

.

Figure 1 Location of the parcels for study

Parcel C

1.5ha

Parcel D

1.5ha

Parcel A

1.5ha

0 100m

Moseushi town

0 100km


Table 1 Cultivation processes and the dates each process was applied (2011)

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."

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

Figure 2 Irrigation and drainage system for parcels

Drainage conduit Feeder road (5m wide)

Irrigation control unit

Sluice for groundwater level adjustment

Surface drain outlet

Water distribution conduit

Drainage conduit

Legend

Water distribution conduit

Underdrainage conduit


Conduit

Faucet

Bulkhead (Height-

adjustable)

Water pipe

Underdrainagepipe

Water supply to soil

Concrete mass

Movable slit for controlling the groundwater level in the parcel

Water flow

Sluice for groundwater level adjustment

Groundwater level in the parcel

Parcel surface

Plan view of the parcels

Parcels Parcel A Parcel C Parcel D

Cultivation technique Transplantingculture

Direct sowing in submerged fields

Direct sowing in submerged fields

Soil puddling May. 10-24 No operation No operationInitial water intake No operation May 29 May 29

Transplanting/Sowing May 29 May 30 May 30Drying up of paddy fields Jun. 26 Jun. 29 Jun. 29Re-submerging of paddy fields

Jul. 4-5 Jul. 4-5 Jul. 4-5

Water management for protein control

Jul. 4 - Aug. 12 Aug. 10-24 Aug. 10-24

Harvesting Sep. 10-24 Oct. 12 Oct. 12


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.

Table 2 Periods defined for the purpose of calculating the quantity of irrigation water

Period Description

I Transplanting culture: From the time for water intake for 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

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 through the end of the irrigation period (Aug. 31)

Figure 3 Observation points in a parcel

1/4

1/4

1/4

1/4

Underdrainage pipes

Observation points for ponding depth

Observation points for groundwater level


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

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)

Parcel A

Parcel C

Parcel D

Sluice left open

SurfaceBoth Both Under-ground

Sluice left open

Surface

Water intake for soil puddling


Re-submerging after drying up

paddy fields

Initial water intake


Re-submerging after drying up paddy fields

Underground Under-ground

Both

Sluice left open


Sluice left open

Water intake

Water intakeShallow water management

Underground Underground Both

Sluice left open Sluice left

open

Water intake

Water intake

Shallow water management


Underground

Averagegroundwater level

Average depth ofponding

Height of paddyfield surface

Prec

ipita

tion

(mm

/d)

Gro

und

wat

er ta

ble

(cm

)

Pond

ing d

epth

(cm

)

Gro

und

wat

er ta

ble

(cm

)G

roun

d w

ater

tabl

e (c

m)

Pond

ing d

epth

(cm

)Po

nding

dep

th (

cm)


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

Figure 5 Changes in the groundwater level and the depth of ponding when the sluice is open

-50

-40

-30

-20

-10

0

10-10

0

10

20

30

40

506/29 0:00 6/29 12:00 6/30 0:00 6/30 12:00 7/1 0:00

Pond

ing

dept

h（cm

）

Grou

nd w

ater

tabl

e（cm

） Parcel C （Without soil puddling）

-50

-40

-30

-20

-10

0

10-10

0

10

20

30

40

507/11 0:00 7/11 12:00 7/12 0:00 7/12 12:00 7/13 0:00

Pond

ing

dept

h （cm

）

Grou

nd w

ater

tabl

e（cm

）Parcel A （With soil puddling）

Ground water tablePonding depthHeight of paddyfield surface

Ground water tablePonding depthHeight of paddyfield surface


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).

Table 3 Amount of water intake by period

Parcel Period I Period II Period III Period IVParcel A 136.2mm 73.7mm 114.0mm 158.9mmParcel C 45.0mm 365.7mm 249.1mm 253.7mmParcel D 47.8mm 379.6mm 316.1mm －


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.

Table 4 Quantity and intervals: Water intake for various farms Cultivation

methodFarm operation Days between water

intake operationsDuration of each

water intake operation

H

Quantity of water intakeL/s/ha

Requireddepth of

watermm

Transplantingculture

Rough soil puddling － 18.2 9.0 58.8


Finishing soil puddling The day after rough soil puddling

20.3 6.1 44.9


Transplanting 3 days after finishing soil puddling

21.2 3.9 29.6

Direct sowing culture

Initial water intake － 10.5 12.3 46.4

Direct sowing culture

Shallow water management 6 14.9 22.5 93.0

Both culturetypes

Middle stage irrigation 5 14.8 5.9 34.7

Both culturetypes


－ 31.4 15.2 112.9

Both culturetypes

Water management for protein control (average)

2 40.7 2.7 35.5

Both culturetypes

Water management for protein control (maximum)

2 35.7 5.4 73.8

Figure 6 Rate and duration of water intake per water intake

0

5

10

15

05/21 05/31 06/10 06/20 06/30 07/10 07/20 07/30 08/09 08/19 08/29

A 圃場

取水

強度

(mm

/h) 代かき

移植低タンパク対策

0

5

10

15

05/21 05/31 06/10 06/20 06/30 07/10 07/20 07/30 08/09 08/19 08/29

C 圃場

取水

強度

(mm

/h)

初期入水

浅水管理

低タンパク対策

0

5

10

15

05/21 05/31 06/10 06/20 06/30 07/10 07/20 07/30 08/09 08/19 08/29

D 圃場

取水

強度

(mm

/h)

初期入水

浅水管理

0

20

40

60

80

100

05/21 05/31 06/10 06/20 06/30 07/10 07/20 07/30 08/09 08/19 08/29

日降

水量

(mm

)


Parcel A

Parcel C

Parcel D


Puddling

Transplanting

Wat

er In

take

rate(

mm

/h)

Wat

er In

take

rate(

mm

/h)

Wat

er In

take

rate(

mm

/h)

Prec

ipita

tion（

mm

/d）






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

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

Figure 8 Parcels of paddy fields used for the simulations

No.10 No.9 No.8 No.7

No.6 No.5 No.4 No.3 No.2 No.1

No.11A 2.4haH 37.36m

A 2.4haH 37.36m

A 2.4haH 37.55m

A 2.1haH 37.05m

A 2.1haH 37.05m

A 2.1haH 37.36m

A 2.3haH 37.36m

A 2.4haH 37.60m

A 4.8haH 37.14m

A 2.4haH 37.55m

A 2.4haH 37.60m

main canal (open channel)

branch pipeline

branch pipeline bran

ch p

ipel

ine

branch pipeline


Figure 7 Flow of calculation in the simulation

Input data on conduits and faucet opening

Calculate the flow rate every 30 minutes

Calculate the quantity of water supplied to a parcel every 30 minutes

Calculate the cumulative quantity of water intake at each parcel

Are calculations for 24 hours finished ?

Start

End

Turn off the faucet for each parcel where water intake is completed

No

Yes


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

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 Assumed water intake patterns for the eight parcels

Day in the 15-day period

Cultivation method 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Direct sowing culture F.I. S.W.M. S.W.M.




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 1

Day in the 15-day period

Cultivation method 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15









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

Water intake pattern 2

F.I.: first intake, R.P.: rough puddling, F.P.: finish puddling, S.W.M : shallow water management, M.S.I.: middle stage irrigation


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)





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: [email protected]

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

C-03


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.

Dojo loach, Misgurnus anguillicaudatus Kara-dojo loach, Paramisgurnus dabryanus

10mm

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.

Third barbell length of the righton the upper jaw, BL

Standard length, SL

Venter

Vertebral edge

Head length, HL Body depth, BD Caudal peduncle depth, CPD

Caudal peduncle length, CPL

Right side

Figure 2. Six body parts measured in this study


2. Materials and Methods 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

Body parts Dojo (n 1) = 312) Kara-dojo (n = 72)

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).

14 18 22 25

Dojo18.1 ± 1.714.1 - 24.7

Kara-dojo18.9 ± 1.915.1 - 23.0

p < 0.01

HLSL%

0

50

100

0

10

20

4 10 16 20

Dojo13.6 ± 2.44.3 - 19.3

Kara-dojo14.4 ± 1.611.4 - 18.9

p < 0.01

CPLSL%

0

40

80

0

10

20

5 9 13 18

Dojo12.7 ± 1.45.4 - 16.7

Kara-dojo15.6 ± 1.311.2 - 18.0

p < 0.01

BDSL%

0

50

100

0

20

40

1 5 9 13

Dojo5.6 ± 1.21.8 - 9.0

Kara-dojo9.7 ± 1.26.7 - 12.6

p < 0.01

BLSL%

0

60

120

0

20

7 10 13 16

Dojo10.0 ± 1.07.2 - 14.7

Kara-dojo12.1 ± 1.59.1 - 15.3

p < 0.01

CPDSL%

0

60

120

0

10

20

SpeciesAverage ± s.d.Range (min – max)Result of t-test

Num

ber o

f the

Doj

o sp

ecim

ens (

n=

312,

)

Num

ber o

f the

kar

a-do

jo sp

ecim

ens (

n=

72,

)

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.

-0.011-0.094

-0.063-0.376**

-0.419**-0.479**

0.199**0.358**

0.706**0.733**

0.166**0.254**

0.503**0.103**

0.222**0.391**

0.416**-0.050

-0.015-0.194

HLSL%

BDSL%

CPLSL%

CPDSL%

Upper: correlation coefficient of the Dojo (n = 312)Lower: correlation coefficient of the Kara-dojo (n = 72)**: p < 0.01Line: regression line

BLSL%

CPLSL%

BDSL%

CPDSL%

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

Y <= 0: the Dojo, Y > 0: the Kara-dojo

Significant

probability

Correlation

ratio Correct discrimination %

Y = 0.241 CPDSL% ** + 0.721 BLSL% ** - 8.140

(0.273) 1) (0.884) ＜0.01 0.804

Total: 95.3 (366 2) / 384 3))

Dojo: 95.5 (298 / 312)

Kara-dojo: 94.4 (68 / 72)

Note) ** p < 0.01, 1) standardized coefficient, 2) the number of correct discrimination, 3) 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.

0

10

20

30 Correct discrimination (number of specimens)Dojo -> Dojo (298)Kara-dojo -> Kara-dojo (68)

Incorrect discriminationDojo -> Kara-dojo (14)Kara-dojo -> Dojo (4)

-5 -4 -3 -2 -1 0 1 2 3 4Score of discriminant function, Y

Num

ber o

f spe

cim

ens

Range of incorrect discrimination(-0.467 - 1.217)

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.





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: [email protected], [email protected] **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

C-07


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. Materials and Methods 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.

Dry

Medium

Wet

Saturated

0.0

1.0

2.0

3.0

4.0

5.0

Volumetric water content (cm3/cm3)

pF

0.0 1.0

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:

∑ +=

P)(IYWP (1)

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:

T)Rs,SM4,SM3,SM2,f(SM1,Y = (2)

where, Y is yield (ton/ha), Rs is total solar radiation (MJ/m2/season), T is average air temperature (oC), 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:

WPbYaF 21 += (3)

Maximize F, subject to:

SMmax SM4SM3,SM2,SM1,SMmin ≤≤ (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. Results and Discussion

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

Seasons Precipitation

(mm)

Temperature (oC) Solar radiation (MJ/m2/season) Minimum Average Maximum

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.

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.

y = 1.00x R² = 0.93

2

4

6

8

10

12

2 4 6 8 10 12 Yield model (ton/ha)

Obs

erve

d Yi

eld

(to

n/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.

Fig. 6 Evolution curves in searching for a maximal value of fitness function

y = 0.01x + 0.49 R² = 0.82

0.50

0.53

0.55

0.58

0.60

0.63

0.65

0 2 4 6 8 10 12 14

Season 1 Season 2 Season 3 So

il m

oist

ure

(cm

3 /cm

3 )

Yield (ton/ha)

Initial

Saturated border

y = 0.01x + 0.51 R² = 0.65

0.50

0.53

0.55

0.58

0.60

0.63

0.65

0 2 4 6 8 10 12 14

Season 1

Season 2

Season 3

Soil

moi

stur

e (c

m3 /c

m3 )

Yield (ton/ha)

Crop development

Saturated border

y = -0.01x + 0.58 R² = 0.27

0.35

0.40

0.45

0.50

0.55

0.60

0.65

0 2 4 6 8 10 12 14

Season 1

Season 2

Season 3

Soil

moi

stur

e (c

m3 /c

m3 )

Yield (ton/ha)

Mid-season

Field capacity border

y = -0.02x + 0.61 R² = 0.44

0.30

0.35

0.40

0.45

0.50

0.55

0.60

0 2 4 6 8 10 12 14

Season 1

Season 2

Season 3

Soil

moi

stur

e (c

m3 /c

m3 )

Yield (ton/ha)

Late season

Field capacity border

0.10

0.15

0.20

0.25

0.30

0 10 20 30 40 50 60 70 80 90 100

Max fitness

Ave fitness

Generation numbers

Fitn

ess


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.

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.

0.300

0.400

0.500

0.600

0.700

0 10 20 30 40 50 60 70 80 90 100

SM1 SM2 SM3 SM4

Generation numbers

Soil

moi

stur

e (c

m3 /

cm3 )

Wet

M

ediu

m

Dry


Table 3 Optimal soil moisture level in each growth stage and its comparison to the irrigation regimes in the third season

Components

Irrigation regimes GA model

Plot 1 Plot 2 Plot 3 Plot 4 Optimal Regime Level

Soil moisture (cm3/cm3) 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





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)[email protected]

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

D-01


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)

IPA

After project(2011-2013)

Survey lot

200mOPAIPAOPA

*

* 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 Apr. 14th, 28th, May 9th ○ ○

2009 5 Mar. 28th, Apr. 10th, 20th, 27th, May 7th ○ ○ during 2010 3 Mar. 12th, 30th, May 1st × ○ after 2011 6 Mar. 3rd, Apr. 5th, 14th, 26th, May 6th, 13th ○ ○

2012 7 Feb. 21st, Mar. 27th, Apr. 2nd, 16th, 26th, May 2nd, 11th ○ ○ 2013 8 Mar. 18th, 28th, Apr. 5th, 17th, 25th, May 2nd, 14th, 29th ○ ○


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

2011 2012

200m

Egg-mass number

01～ 56～10

11～

No data

2013

Fig. 4 Distribution of egg-masses in 2008-2013


0

100

200

2008 2009 2010 2011 2012 2013

Den

sity

(num

ber /

km)

OPA

IPA

Project

Year

Fig. 5 Changes of egg-mass density in IPA and OPA

0.00

0.20

0.40

0.60

2008 2009 2010 2011 2012 2013

Den

sity

ratio

(IPA

/ O

PA)

Project

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)





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 : [email protected]; +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,’

D-02



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. Materials and methods 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 condition

(biomass)

Trophic interaction (preference, %)

Copepod

Cladoceran

Rotifer

Minnow

Bluegill Carp Bass

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

Plants Phytoplankton

Diatom 0.553 mg/L dry 18.4

Greens 0.481 mg/L dry 18.4 5.6

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

Invertebrates Susp Feeders

Copepods 0.017 mg/L dry 14.3 100.0 11.1 1.5

Cladocerans 0.069 mg/L dry 14.3 11.1

Rotifers 0.011 mg/L dry

Fish

Small forage Minnow 3.13 g/m2 dry 10.6

Large forage Bluegill 2.25 g/m2 dry 7.3

Large bottom Carp 0.58 g/m2 dry 14.7

Large game Largemouth Bass 0.30 g/m2 dry

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

Aquatic life Exp. Time

(hr) Endpoint

Alachlor (µg/L)

Paraquat (µg/L)

Copper sulfate (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

Parameter Description Unit Model range

Possible range

Initial value

Final valuse

INFILT Index to infiltration capacity in./hr 0.0001

~ 100.0 0.001

~ 0.5 0.16 0.10 ~ 0.27

KVARY Variable groundwater recession 1/in. 0.0 ~ 5.0 0.0 ~ 5.0 0.0 0.00 ~ 0.8

AGWRC Base groundwater recession none 0.001

~ 0.999 0.85

~ 0.999 0.98 0.92 ~ 0.98

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

Parameter Definition Unit Model range

Initial value

Final value

KBOD20 Unit BOD decay rate at 20℃ 1/hr 1.0E-30 ~ none 0.004 0.004

~ 0.067

KODSET BOD settling rate ft/hr 0 ~ none 0.027 0.011 ~ 0.027

REAK Empirical constant in the equation used to calculate the reaeration coefficient 1/hr 1.0E-30

~ none 0.2 0.05 ~ 0.2

BRBOD Base release rate of BOD materials mg/m2 0.0001 ~ none 0.001 0.001

~ 150.0

CVBO Conversion from milligrams biomass to milligrams oxygen mg/mg 1.0 ~ 5.0 1.63 1.63

CVBPC Conversion from biomass expressed as phosphorus to carbon moles/mol 50 ~ 200 106 106

CVBPN Conversion from biomass expressed as phosphorus to nitrogen moles/mol 10 ~ 50 16 16

BPCNTC Percentage of biomass which is carbon none 10 ~ 100 49 49

BRNIT The benthal release rates of ammonia under aerobic and anaerobic condition mg/m2 0.0

~ none 0.0 0.0 ~ 95

KTAM20 Nitrification rates of ammonia at 20℃ 1/hr 0.001 ~ none 0.045 0.045

KNO220 Nitrification rates of nitrate at 20℃ 1/hr 0.001 ~ none 0.008 0.008

KNO320 Denitrification rates of nitrate at 20℃ 1/hr 0.001 ~ none 0.008 0.008

MALGR Algal growth rate 1/hr 0.001 ~ none 0.0085 0.085

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:

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.

http://webbook.me.go.kr/DLi-File/086/5508730.pdf

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: NIER, National Institute of Environmental Research., 2006, Survey on Environment and Ecosystem of

Lakes in the Han river.

http://www.hanriver.re.kr/App_JSP_Page/tb_resc_info.jsp?s_resc_grp_cd=1

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.

http://webbook.me.go.kr/DLi-File/086/5508730.pdf�

http://www.hanriver.re.kr/App_JSP_Page/tb_resc_info.jsp?s_resc_grp_cd=1�


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.





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

D-05


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. Materials and Methods 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 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.

Fig. 1 Study area


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 3.1 Water and Nitrogen balance The results showed in Table1. Based on these water and nitrogen balance, virtually averaged water quality was estimated in Table1. The result of nitrogen water quality was 1.3 mg/l that was lower than 3.8 mg/l in averaged river water quality. This is the influence of denitrification in paddy fields area. Also, discharge volume was 0.33×106m3 that was assumed to be available to around 40 ha for 20 mm/day consumption in paddy fields. This area was 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.

0

5

10

15

20

25

30

35

40

450

10

20

30

40

50

60

7/26 8/26 9/26 10/26 11/26 12/26

日降雨量比流量

Fig.2 Rainfall and discharge（2012/7/26～12/31）

Rainfal Discharge

0

10

20

30

40

50

60

0

2

4

6

8

10

12

7/18 8/18 9/18 10/18 11/18 12/18

No.25(N) 排水路比流量mg/L mm/d

Fig.3 Water quality (T-N) in discharge and drainage channel

Qin Qout Ir D Rc BG Rain ET

灌漑期

5.70 7.46 0.23 0.33 0.98 2.64 0.93 3.83

非灌漑期

22.17 28.00 0.01 2.06 8.06 11.84 13.94 5.03

Lin Lout L(Ir) L(Rc) L(D) L(BG) L(dSw) L(P)

灌漑期 19.16 16.20 0.77 3.31 0.64 0.25 3.41 0.23

非灌漑期

79.66 71.58 0.05 28.95 0.00 17.45 15.01 3.46

Cin Cout C(Ir) C(Rc) C(P) C(D) C(BG)

灌漑期 3.36 2.17 3.36 3.36 0.25 1.95 0.10

非灌漑期

3.59 2.56 3.59 3.59 0.25 0.00 1.47

Table 1 Water Balance (up)，Nitrogen Balance (middle)，average nitrogen

(d )

Irrigation

Irrigation

Irrigation

Non-Irrigation

Non-Irrigation

Non-Irrigation


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).





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:[email protected], 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: [email protected]

***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: [email protected]

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

D-07





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. Materials and methods 2.1 Survey site

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.

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 fields are divided into 5 areas based on water flow from 5 small rivers.

The beneficial area of River A (hereinafter referred to as ‘Area A’) was developed in the 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

Vientiane Province

Nam Ngum Dam

5km

Nam Ngum Dam N Village

Feauang


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.

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

Riv

er A

Riv

er B

Riv

er C

Riv

er E

Riv

er D


Fig. 3 Numbering of survey fields

Owned by farmers of the other village The Fields numbered with red character are sampling fields The fields which can be taken water from canal

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.

Rectangular weirs with automatic water-level gauges (On-set HOBO U-20) were installed at 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. 3. Results and discussion 3.1 Precipitation in 2012

Table 1 shows monthly precipitation in Feuang District in 2011, 2012, and the average of 6 years (from 2007 to 2012). Average an-nual precipitation was 2,380mm. Precipitation exceeded 2800 mm in 2011, a year in which there was heavy rain and extensive flooding in the Indo-China area. In 2012, there was less rainfall than both the average and the 2011 rainfall. Cumulative precipitation before transplanting from March to June in 2012 was lower, and it was half of that in 2011. Less precipitation occurred in 2012, resulting in poorer conditions for nursery establishment and transplanting compared to 2011 and the average. There was heavy rain in July and August, consistent with a typical year.

Table 1 Monthly precipitation in Feuang District

2011 2012 Avg. (2007-2012)

Jan 1.3 5.3 19.9 Feb 6.1 10.4 6.4 Mar 152.7 21.2 54.0 Apr 171.3 102.3 119.0 May 476.3 326.2 337.0 Jun 568.4 208.8 361.2 Jul 285.9 418.7 355.4 Aug 435.5 511.5 475.6 Sep 559.3 348.8 406.8 Oct 181.7 77.3 194.5 Nov 39.5 92.3 41.0 Dec 3.4 5.7 7.0 Total 2,881.4 2,128.5 2,377.8

Based on the data observed at Feuang station by DoMH.

Area A Area B


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,

they were counted two or more times. The nurseries established outside of Areas A and B were not counted. Of 53 survey fields, transplanting was not completed in 5 fields by 4 August and in 1 field, rice was not planted in 2012.

Establishment of nurseries began in the 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 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-

Fig. 3 Time of farming activities and daily precipitation from 28 May to 4 August

Fig. 4 Completion time of transplanting

~ 28July 28 July ~ Direct seeding

~ 07 July ~ 14 July ~ 21 July


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.

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

Table 2 Rice yield in the sampling fields

Field No Yield

(ton/ha) Std. dv. Note

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%


than 3 days. In low-elevation fields, such situations might damage rice growth and productivity. Field 51 was under watery conditions, because it was located near a swamp. A local, long-stem rice variety (Makkua) was planted in this field. It was 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 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 Field type

Average (ton/ha) Std. dv. Note

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 Area


Area A 4.09 1.39 n=72 Area B 3.74 1.20 n=18 Nosignificant difference by t-test Table 5 Average rice yield of each variety of rice Variety


Tia khao 4.37 1.13 n=18 Takkied 4.02 1.39 n=15 Tia daeng 3.83 1.17 n=33 The others 4.02 1.72 n=24 Nosignificant difference by Tukey-Kramer HSD test

Fig. 6 Water discharge from River A and B

Table 6 Average rice yield of difference time of transplanting Translanting


2 – 8 July 2013 5.00 2.51 n=9 9 – 15 July 203 3.97 1.39 n=51 16 – 22 July 2013 3.94 1.17 n=21 Later than 23 July 3.51 1.72 n=9 Nosignificant difference by Tukey-Kramer HSD test

Fig. 5 Map of rice yield in the survey area

3 ton < 2 ton <

6 ton < 5 ton < 4 ton <


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_progress_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

http://www.fao.org/fileadmin/templates/ess/documents/food_security_statistics/monitoring_progress_by_country_2003-2005/LaoPDR_e.pdf�

http://www.fao.org/fileadmin/templates/ess/documents/food_security_statistics/monitoring_progress_by_country_2003-2005/LaoPDR_e.pdf�


PAWEES 2013 (12TH) INTERNATIONAL

CONFERENCE ON AGRICULTURAL WATER AND RURAL ENVIRONMENT FOR

THE FUTURE RAMADA PLAZA HOTEL, Cheongju, KOREA


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.

D-08


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). 3. Materials and Methods 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.

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.

outlet

inlet drainage channel

→irrigation channel


3.1.2 Questionnaire of rice farming procedure

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.

0

50

100

150

200

250

300

6/12～6/18 6/19～6/25 6/26～7/2 7/3～7/9 7/10～7/16

総羽

化殻

数

調査期間

アキアカネ

ノシメトンボ

ナツアカネ

アカネ属以外

不明

Sympetrum frequens Sympetrum darwinianum Sympetrum infuscatum Except Sympetrum genus Undistinguishable

Study period

Num

ber of emergence

Figure 3 Time series of dragonflies’ emergence in Niigata Prefecture, 2011

Lower lip

Side spine

9th node 8th node

Figure 2 Points to identify dragonfly husks Side spine of 8th node straight toward 80% of 9th node.


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.

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.

Figure 4 Manifestation rate of emergence husks of red-dragonflies in Niigata Prefecture, 2011.

Sympetrum frequens Sympetrum darwinianum Sympetrum infuscatum Non-Sympetrum genus Undistinguishable

0

2

4

6

8

10

12

14

16

中干し無し中干し有り

平均羽化殻数（標準偏差）

Non-midterm drainage Midterm drainage

Num

ber of emergence

(standard deviation)

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.

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.

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

0

2

4

6

8

10

12

14

16

18

有機除草剤のみ除草剤・箱施用剤

平均

羽化殻

数(標

準偏

差)

Organic Herbicide Seedbox insecticide + Herbicide

Num

ber of emergence


Figure 6 Effect of herbicide and/or seedbox insecticide to dragonfly emergence in Niigata Prefecture, 2011.

0

2

4

6

8

10

12

14

16

18

20

ネオニコチノイド系

（フィプロニル含む）N=9

ヘテロ系

N=8

ネライストキシン系

N=3

平均羽化殻数

Neonicotinoid Hetero Nereistoxin

Figure 7 Relationship between branches of the seed-box insecticide and dragonfly emergence in Niigata Prefecture, 2012.

Num

ber of emergence



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





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)

E-01


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.

2. Materials and Methods 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).

0 10 20 30 40 50 60 70 80 90 100110120130140150160

Soil M

oistu

re (%

)

16

18

20

22

24

26

Redo

x (M

v)

100

200

300

400

500

600

Rainf

all (m

m)

0

20

40

60

80

100

120

June July Aug. Sept. Oct. Nov.

Germination

Transplanting

Water Application

Water Application

Harvesting

Flowering time

Fig. 1. Environmental condition in the aerobic field during the growing period in wet season 2012 at

Kasetsart University, Kamphaeng Saen, Thailand.


3. Results and discussion 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.

Fig. 2. Grain yield (kg/ha) of selected rice varieties under SRI (aerobic) (blue) and flooded (red) condition.

0.0

100.0

200.0

300.0

400.0

500.0

600.0

IR29

IR40

IR42

IR70

IR72

PSB

Rc3

0PS

BR

c9PS

BR

c70

Ket

an L

um

bu

IRG

C35

747

( …

SPTC

9903

9SP

TC99

052

( PTT

)SP

TC04

072

( PTT

)SP

TC04

069

SPTC

0407

0 ( P

TT )

Bu

eogi

( PT

T )

PSL

0003

4-50

-2-1

-8PS

L 00

034-

50-2

-1-3

PSL

0005

04-5

4-1-

2-2

WS4

8 PT

T1W

S48

SPR

2W

S48

SPR

90FL

496

RD

15Pi

tsan

ulo

kPi

nka

set 3

Sin

tan

ur (

SRI)

RG

DU

073

36-4

3-10

-24

RG

DU

073

45-4

-24-

1R

GD

U 0

7345

-4-2

4-41

CH

AIN

T1PT

T1IR

5751

4SU

RIN

1Su

ph

anb

uri

60

SRI (Aerobic)Flooded (Anaerobic)


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 spacingon 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.





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 University, Chuncheon, Gangwondo, South Korea; [email protected]

**Researcher, Agro-climatology Lab., Rural Development Administration, Suwon, Gyeonggido, South Korea; [email protected] ***Post-doctoral researcher, Department of Regional Infrastructures Engineering, Kangwon National University, Chuncheon, Gangwondo, South Korea; [email protected]

****Researcher, Department of Regional Infrastructures Engineering, Kangwon National University, Chuncheon, Gangwondo, South Korea; [email protected] *****Director, Agro-climatology Lab., Rural Development Administration, Suwon, Gyeonggido, South Korea; [email protected] ******Researcher, Research Group on Rural Community Development, Rural Research Institute, Ansan, Gyeonggido, South Korea; [email protected] *Corresponding author: [email protected], ☎+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

E-02










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.

2. Materials and methods

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. Results and discussions

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

Interval

Irrigation supply (㎥)

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

Concen-

tration

(mg/l)

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

CS 2.9b 32.8b 5.6b 15.3b 4.85ab 0.32b

SRI 2.55b 33.80b 5.75b 16.15b 4.13b 0.34b

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%

Load

(kg/ha)

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

CS 34.8b 562.6b 55.2b 168.0b 64.3b 3.4b

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

Treatment Emissions (kg/ha) CO2 ton/ha

equivalent

Index

(%) CH4 N2O

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

Treatment Measured date (2011)

27 May 1 Jun 8 Jun 22 Jun 1 Jul 20 Jul 5 Aug

No. of

tillers/hill

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

SRI-30/2 13 16 18 35 47 52 48

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

Plant

height

(cm)

CT 16 20 26 41 50 85 104

CD 19 23 28 44 52 82 106

CS 16 20 27 47 62 86 108

SRI-30/1 18 24 31 48 63 86 106

SRI-30/2 20 23 31 49 63 89 110

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

Treatment Weight/

hill (g)

Panicles/

hill

Grains/

panicle

Effective

grains/panicle

(%)

Weight/

1000

grains(g)

Yield (kg/10a) Yield index

Polished

rice

Head

rice

Polish

ed rice

Head

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/

http://www.cropscience.org.au/�





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

E-04


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.

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.

Shihmen Reservoir

Taoyuan County

Taipei City

Taoyuan Canal


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 domestic balance of

rice supply and demand

The plantation areas of rice was declined from 36.4*104 ha in 1997 to 26.9*104 ha in 2005, and 25.4*104 ha in 2011. The policy objectives of maintaining domestic balance of rice supply and demand was achieved.

Stabilize the market prices

The annual average price per kilogram of Japonica Rice was $18.68 for the previous 3 years by 2001, and increased to $22.12 in 2009. It was reached the most ideal state for adjusting the industrial structure of rice and stabilized the market prices effectively.

Ensure the income of farmers

The average farmer earning of 1st crop rice was $55,772 for the previous five years by 2001, and increased to $60,407 for the last five years. It was ensured the farmers income effectively.

Reduce the domestic support for

agriculture (AMS)

Reduced the AMS of the rice of public grain from 4.82 billion in 2001 to 2.77 billion in 2008. We compared the AMS in 2008 to the base year of 2001, the domestic AMS has dropped from 17.71 billion to 4 billion, a drop of 77%, reaching reduction goal.

Table 2 The Comparison of rice supply and demand of Taiwan for joining the WTO

year Area of Rice

of Taiwan (103 ha)

Population of Taiwan

(103 )

Total Consumption of Taiwan

(103 ton)

Total Supply of Taiwan (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

year

1st crop-term 2nd crop-term Area

of rice (103 ha)

Reservoir

supply (106 m3)

River supply

(106 m3)

Irrigation depth (mm)

Irrigation loss

(106 m3)

Field irrigation depth (mm)

Area of rice (103 ha)

Reservoir

supply (106 m3)

River supply

(106 m3)

Irrigation depth (mm)

Irrigation loss

(106 m3)

Field irrigation depth (mm)

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.

Fig. 2 Irrigation Water Use and Rice Area of TIA

0

20

40

60

80

100

0

50

100

150

200

1995

19

96

1997

19

98

1999

20

00

2001

20

02

2003

20

04

2005

20

06

2007

20

08

2009

20

10

2011

20

12

Reservoir Supply

year

Area (103 ha)

Wat

er S

uppl

y (1

06 m3 )


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

year

Planned irrigation

water demand (106 m3)

Reservoir supply

(106 m3)

Support for Water Company

(106 m3)

Real irrigation water

(106 m3)

Real utilization of

irrigation water

Water Saving (106 m3)

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

Reservoir Supply

Rice Area & FID by Reservoir & River

Supply

Rice Area & FID by River Supply

Total Rice Area & FID

Average 1st Crop-term in 2002~2012

5,669 ha 1,371 mm

3,356 ha 982 mm

470 ha 5,847 mm

9,495 ha 1,455 mm

1st Crop-term in 2013

7,898 ha 893 mm

4,037 ha 795 mm

592 ha 4,957 mm

12,527 ha 1,054 mm

Ratio 139% 65%

120% 81%

126% 85%

132% 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.





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.

E-05


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

and distributary (secondary) canals (MC and DY canals)

Starting irrigation service

Stage IV Construction of minor (tertiary) canals On-farm irrigation facilities Stage V Construction of water courses (WC) and

organizing water user groups (WUGs) On-farm irrigation facilities and 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.

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

Table 2. Water supply system in Baingda Project

DY canal Number of Minor canals for water

supply to left

Number of Minor canals for water supply to right

DY-1 7 Nos. 5 Nos. DY-2 2 Nos. 11Nos. DY-3 2 Nos. 5 Nos. DY-4 3 Nos. 1 No. DY – Distributary , Minor canal – Tertiary canal

Figure 1. Layout of canals in Baingda Project

Baingda weir

Baingda dam

DY-1

DY-2

DY-3

DY-4

Main canal

Irrigable area


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).

0

200

400

600

800

1000

1200

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Mon

thly

rai

nfal

l (m

m)

Month

2010

2011

2012

Figure 2. Monthly rainfall in the project area, Daik-U town, (2010, 2011 & 2012)

0

500

1000

1500

2000

2500

3000

2010-11 2011-12 2012-13

Irri

gate

d ar

ea (h

a)

Irrigation season (Year)

0

1000

2000

3000

4000

5000

6000

DY-3, Minor-4 area

Project' total area

Cul

tivat

ed a

rea

(ha)

Location

Non cultivation

Upland crop

Paddy

Figure 3. Irrigated area of the project area Figure 4. Cultivated area in the dry season, 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

Figure 5. Distribution of on-farm plots and the cultivated area of Minor-4 canal area of DY-3 canal, 2012-13

Figure 6. Distribution of plots in elevation in Minor-4 canal area of DY-3 canal

Figure 7. Water release from the reservoir for irrigation

0.00

50.00

100.00

150.00

200.00

250.00

300.00

350.00

400.00

450.00

500.00

10-N

ov

20-N

ov

30-N

ov

10-D

ec

20-D

ec

30-D

ec

10-J

an

20-J

an

30-J

an

10-F

eb

20-F

eb

30-F

eb

10-M

ar

20-M

ar

30-M

ar

10-A

pr

20-A

pr

20-A

pr

10-M

ay

20-M

ay

30-M

ay

Wat

er st

orag

e in

rese

rvoi

r (M

CM

)

Date

2009-102010-112011-122012-13Full storageLow storage


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.

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.

Figure 8. Water supply in canals, 2012-13 season

0.00

0.50

1.00

1.50

2.00

2.50

25-D

ec-1

2

1-Ja

n-13

8-Ja

n-13

15-J

an-1

3

22-J

an-1

3

29-J

an-1

3

5-Fe

b-13

12-F

eb-1

3

19-F

eb-1

3

26-F

eb-1

3

5-M

ar-1

3

12-M

ar-1

3

19-M

ar-1

3

26-M

ar-1

3

2-A

pr-1

3

9-A

pr-1

3

16-A

pr-1

3

23-A

pr-1

3

30-A

pr-1

3

7-M

ay-1

3

14-M

ay-1

3

21-M

ay-1

3

Can

al w

ater

sup

ply

in d

epth

(m

)

Irrigation period (Date)

Main canal DY-3 canal Minor-4 cnal


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. 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

Total area (Project whole area)

DY-3 area

Minor-4 area

Figure 9. Layout of water balancing study in Baingda Project

F & NF

Chairman

Secretary

Members Village (i)

Members Village (ii)

Members Village

F & NF F & NF

Figure 10 The organization of a Village Tract in administrative body

F - Farmers NF - Non-farmers


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





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:[email protected])

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.

E-06


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.


2. Materials and methods

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 control

efficiency (%) at 60

DAT

No of tillers (m-2)

No of filled grains

panicle-1

Test weight (g)

Yield (tone ha-1)

Net return (Rupees/Rs

ha-1)

Net return

(Dollar/$ ha-1)

Y1 Y2 Y1 Y2 Y1 Y2 Y1 Y2 Y1 Y2 Y1 Y2 Y1 Y2

Brown manuring +

Recommended N

77.38 80.73 406.66 405.00 154.34 144.55 22.21 21.43 5.18 5.41 30889 39442 498 636

Brown manuring +

Vermicompost with N through

LCC

72.02 74.18 430.00 420.33 152.31 141.59 21.46 21.22 5.50 5.76 27191 36331 439 586

Brown manuring + N through LCC

64.73 66.72 441.66 430.66 151.47 141.51 23.55 21.27 5.22 5.56 31665 41536 511 670

Mulching with maize residue + Recommended

N

61.63 63.73 396.33 388.33 136.20 128.10 20.14 19.66 5.03 5.25 30683 38987 495 629

Mulching with maize residue + Vermicompost with N through

LCC

61.61 63.79 424.66 416.33 134.30 127.94 21.94 18.55 5.20 5.52 24044 33627 388 542

Mulching with maize residue + N through LCC

61.50 63.96 420.33 411.00 135.59 127.72 22.45 18.72 5.13 5.33 31891 40031 514 646

Hand weeding + Recommended

N

100 100 389.33 380.66 158.73 150.11 22.38 21.67 5.22 5.38 24416 32175 394 519

Hand weeding + Vermicompost with N through

LCC

100 100 401.00 390.33 159.00 149.58 23.28 22.31 5.54 5.87 19557 29548 315 477

Hand weeding + N through LCC

100 100 394.66 385.00 161.22 148.62 22.75 21.36 5.32 5.65 25750 35586 415 574

Weedy control + Recommended

N

0.00 0.00 293.66 281.66 94.77 86.39 20.55 19.30 1.99 2.53 -37 9942 -1 160

Weedy control + Vermicompost with N through

LCC

0.00 0.00 290.33 279.33 94.24 86.12 19.40 19.41 2.21 2.85 -5973 5220 -96 84

Weedy control + N through LCC

0.00 0.00 285.33 277.00 92.20 86.09 18.67 19.14 2.07 2.64 180 10463 3 169

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





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

F-02


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

ACCESS1.0 CSIRO (Commonwealth Scientific and Industrial Research Organisation, Australia), and BOM (Bureau of Meteorology, Australia)

BCC-CSM1 Beijing Climate Center, China Meteorological Administration CCSM4 National Center for Atmospheric Research

CNRM-CM5 Centre National de Recherches Meteorologiques / Centre Europeen de Recherche et Formation Avancees en Calcul Scientifique

http://www.kma.go.kr/�


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

HADGEM-ES Met Office Hadley Centre , Instituto Nacional de Pesquisas Espaciais, Korea Meteorological Administration

INMCM4 Institute for Numerical Mathematics IPSL-CM5A Institut Pierre-Simon Laplace

MIROC5 Atmosphere and Ocean Research Institute (The University of Tokyo), National 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 influence 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.): ( )0.5 1 kLAI

d actAGB RUE Rad e−= × ⋅ ⋅ − (1) where Rad (MJ m-2d-1) is daily global solar radiation, 1-e-kLAI 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): max lim ( _ ) ( _ )actRUE RUE T Rad F DVS F= ⋅ ⋅ ⋅ (2) where RUEmax (kg MJ-1) 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);

lim

c b

o b

CT TT T

b c

o b c c

T T T TTT T T T

−−

− − = − −

(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 parameters Value unit Growth parameters Value Unit

Base T before emergence 12 °C Radiation use efficiency 2.4 g MJ-1

Max. T before emergence 40 °C Extinction coefficient for solar

radiation 0.5 -

GDDs to reach emergence 80 °C day Base T for biomass accumulation 12 °C

Base T after emergence 12 °C Optimum T for biomass

accumulation 28 °C

Max T after emergence 40 °C Max T for biomass accumulation 35 °C

Leaf area index at emergence 0.02 m2 m-2 GDDs emergence – flowering 970 °C day Specific leaf area at emergence 27 m2 kg-1

Specific leaf area end tillering 18 m2 kg-1 GDDs flowering – maturity 550 °C day Fraction of biomass partitioned to

leaves at emergence 0.8 -

Development stage at harvest 2.0 - Kc full canopy 1.2 -

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):

1min( , )

t

eff ii

green

WD RWF

Y==∑

(4)

1max(0, )

n

eff ii

blue

WD RWF

Y=

−=∑

(5)

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. Results and discussion 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.

http://dspace.inea.it/�


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.





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

F-03


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. Results and discussions 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.


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

WGLEHRn =−−− (1)

where Rn is the net radiation (W/m2), H is the sensible heat flux (W/m2), LE is the latent heat flux

14

16

18

20

22

24

26

1-Jul 11-Jul 21-Jul 31-Jul 10-Aug 20-Aug 30-Aug

U A B C D E F G

20242832


Air temperature

02468


Discharge at I

Wat

er T

empe

ratu

re (d

eg C

) (

deg

C)

(m

3 /s)


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:

( ) 41 wdn TLSR εσα −+−= (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.

Fig. 5 Calculated heat fluxes at the plot G, 30 July - 2 August 2012

18

20

22

24

26

28

30-Jul0:00

31-Jul0:00

1-Aug0:00

2-Aug0:00

3-Aug0:00

4-Aug0:00

5-Aug0:00

6-Aug0:00

U A B C D E F G

Wat

er T

empe

ratu

re (d

eg C

)

-200

0

200

400

600

800

30-Jul 0:00 31-Jul 0:00 1-Aug 0:00 2-Aug 0:00 3-Aug 0:00

Rn H LE G W

Hea

t flu

x (W

/m2 )


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:

( )Ein

IVEIVin TT

TTQQ−−

= (3)

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.

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)

16

17

18

19

20

0:00 6:00 12:00 18:00 0:000

20

40

60

80

IVEDepth of the drainage channel (right axis)

Depth (cm

)

Wat

er te

mpe

ratu

re (d

eg C

)

13

14

15

16

17

0:00 6:00 12:00 18:00 0:000

20

40

60

80

IVEDepth of the drainage channel (right axis)

Depth (cm

)

Wat

er te

mpe

ratu

re (d

eg C

)


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)

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.

0.00

0.05

0.10

0.15

0.20

0:00 6:00 12:00 18:00 0:000

20

40

60

80

Estimated discharge of return flowDepth of the drainage channel (right axis)

0.00

0.05

0.10

0.15

0.20

0:00 6:00 12:00 18:00 0:000

20

40

60

80

Estimated discharge of return flowDepth of the drainage channel (right axis)

Depth (cm

)

Depth (cm

)

(m

3 /s)

(m

3 /s)

202224262830323436

15-Jul0:00

16-Jul0:00

17-Jul0:00

18-Jul0:00

19-Jul0:00

20-Jul0:00

21-Jul0:00

22-Jul0:00

23-Jul0:00

24-Jul0:00

25-Jul0:00

Drainage water temperature Air temperature

182022242628303234

10-Jun0:00

11-Jun0:00

12-Jun0:00

13-Jun0:00

14-Jun0:00

15-Jun0:00

16-Jun0:00

17-Jun0:00

18-Jun0:00

19-Jun0:00

Drainage water temperature Air temperature

Tem

pera

ture

(deg

C)

Tem

pera

ture

(deg

C)


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

DcW

xT

AQ

tT

ww

ww

ρ=

∂∂

+∂∂

(4)

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.

14182226303438

30-Jul0:00

31-Jul0:00

1-Aug0:00

2-Aug0:00

3-Aug0:00

4-Aug0:00

5-Aug0:00

6-Aug0:00

7-Aug0:00

14182226303438

22-Jun0:00

23-Jun0:00

24-Jun0:00

25-Jun0:00

26-Jun0:00

27-Jun0:00

28-Jun0:00

29-Jun0:00

30-Jun0:00

Air temperature 1 2 3 4 5 6

Tem

pera

ture

(deg

C)

Tem

pera

ture

(deg

C)


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:

G_obsFG_calF_obsin

FG_calF_obsG_obs

)()(

QTTT

TTTQ

ww

wwwR ∆+−

∆+−= (5)

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.

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

0.0E+001.0E-042.0E-043.0E-044.0E-045.0E-046.0E-047.0E-048.0E-04

25-Jul0:00

26-Jul0:00

27-Jul0:00

28-Jul0:00

29-Jul0:00

30-Jul0:00

31-Jul0:00

1-Aug0:00

2-Aug0:00

3-Aug0:00

4-Aug0:00

5-Aug0:00

6-Aug0:00

D - IV IV - E E - F F - G

R

etur

n flo

w ra

te p

er

padd

y fie

ld a

rea

(mm

/s)

18

20

22

24

26

28

30-Jul0:00

31-Jul0:00

1-Aug0:00

2-Aug0:00

3-Aug0:00

4-Aug0:00

5-Aug0:00

6-Aug0:00

7-Aug0:00

A_obs G_obs G_cal (with climatic factor)

Wat

er te

mpe

ratu

re (d

eg C

)


return flow KR as:

)( inin

www

ww TTA

qDc

Wx

TAQ

tT

−+=∂∂

+∂∂

ρ (6)

where qin (m2/s) is the discharge rate of return flow per unit length of channel which can be calculated by the equation:

LAreaKq R ÷×÷×= − 8640010 3in (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.

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

161718192021222324252627

8-Aug0:00

9-Aug0:00

10-Aug0:00

11-Aug0:00

12-Aug0:00

13-Aug0:00

14-Aug0:00

15-Aug0:00

16-Aug0:00

17-Aug0:00

161718192021222324252627

9-Jul0:00

10-Jul0:00

11-Jul0:00

12-Jul0:00

13-Jul0:00

14-Jul0:00

15-Jul0:00

16-Jul0:00

17-Jul0:00

18-Jul0:00

A_obsG_obsG_cal (with climatic factor)G cal (with climatic factor and return flow 13.4mm/d)

Wat

er te

mpe

ratu

re (d

eg C

) W

ater

tem

pera

ture

(deg

C)


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 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

F-05


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. Materials and Methods 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.

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 (Gs)

Cu Cg LL (%)

PL (%)

USCS

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

Mulwang(MW)

Ansung(AS)

Jechon(JC)


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

Case Condition of fill dam Water level Seismic Safety factor This

Study upstream downstream 1

As usual Rapid down None 1.2 1.2 O

2 Filled with water Concern 1.2 1.2 O 3 Usual level Concern 1.15 - O 4 Shortly after

construction Bottom None 1.3 1.3 -

5 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 Embankment 18.48 25 17 7.83 X 10-5

EAP plan of reservoirs. (KRC, 2008)

Core 19.79 14 39 2.61 X 10-7 Soft rock 24.53 35 100 1.00 X 10-9

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


1.659

Distance0 20 40 60 80 100 120 140 160 180

Elev

atio

n

0

10

20

30

40

50

60

1.227

Distance0 20 40 60 80 100 120 140 160 180

Elev

atio

n

0

10

20

30

40

50

60

(a) Rapid down water level at upstream (b) As usual water level at upstream(/w seismic)

1.304

Distance0 20 40 60 80 100 120 140 160 180

Elev

atio

n

0

10

20

30

40

50

60

(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

1.709

Distance0 20 40 60 80 100 120 140 160 180

Elev

atio

n

0

10

20

30

40

50

60


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.

http://search.cnki.com.cn/Search.aspx?q=ISO%2010390:2005�

http://search.cnki.com.cn/Search.aspx?q=Soil%20quality�

http://search.cnki.com.cn/Search.aspx?q=pH�

http://search.cnki.com.cn/Search.aspx?q=Difference%20in%20method�

http://search.cnki.com.cn/Search.aspx?q=ISO%2010390:2005�





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.

F-08


On the other hand, owing to recent rapid development of information communication technology (ICT), it has become possible to instantly transmit information on the amount of available water and on the demand of farmers interactively, making use of Internet and various field sensors (Olalla et al., 2003). 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 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 farmers in regard to their irrigation management have not been scientifically investigated so far. It is 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 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 on a service dominant logic.

2. Methods 2.1 Site description 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.

Fig. 2 Experimental plots

P

Plot 2

Plot 1

: Flow meter a=command area M

M

M

140o E

130o E

40o N The Japan Sea

Study area

Fig. 1 Location of the study area


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. 2.3 Water balance at a plot The daily water balance at each experimental plot was calculated for the whole cultivation period. Because the farmer cultivating Plot 1 mentioned that a lot of lateral seepage between plots occurred, the water balance at a plot was evaluated taking lateral seepage into account, using the equation

I + Pi + R = ET + D + Po + dH (1)

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 (mm/d), ET is the evapotranspiration (mm/d), D is the amount of surface drainage water (mm/d), Po is the sum of the lateral seepage outflow to the adjacent plot and the vertical percolation (mm/d), dH is the changes in the floodwater depth (mm/d). Eq. 1 is rearranged in

(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.

Table 1 Characteristics of the farmers cultivating the selected plots

Plot 1 Plot 2 Involvement in farming Full-time Part-time Farming experience (y) >40 >40 Number of workers (persons)

2 2

Distance from the residence (km)

2.1 2.4

Other information Retired person farming for enjoyment

Table 2 Cultivation calendar Plot 1 Plot 2 Puddling May 23 Apr. 30 Transplanting May 29 May 5 Mid-summer drainage

from Jun. 29 Jun. 14 to Jul.16 Jul. 14

Full drainage Sept. 30 Aug. 19 Harvest Oct. 7 Sept. 16


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.

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.

Date (2012)5/1 6/1 7/1 8/1 9/1 10/1 O

peni

ng le

vel

5/1 6/1 7/1 8/1 9/1 10/1

Dep

th (m

m/d

)

0

40

80

120PrecipitationIrrigationPercolation (Into)

5/1 6/1 7/1 8/1 9/1 10/1

Dep

th (m

m/d

)

0

40

80

120EvapotranspirationPercolation (Out) + Drainage

fullmiddle

littleshut

mid-summerdrainageP T F

(a)

(b)

(c)


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

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.

Date (2012)5/1 6/1 7/1 8/1 9/1 10/1

Ope

ning

leve

l

5/1 6/1 7/1 8/1 9/1 10/1

Dep

th (m

m/d

)

0

100

200

PrecipitationIrrigationPercolation (Into)

5/1 6/1 7/1 8/1 9/1 10/1

Dep

th (m

m/d

)

0

100

200EvapotranspirationPercolation (Out) + Drainage

fullmiddle

littleshut

P T

F

mid-summerdrainage

(a)

(b)

(c)


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.

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.

Table 3 Results from interviews with the cultivating farmers

Questions Answers

The farmer cultivating Plot 1 The farmer cultivating Plot 2

Opinion about outsourcing of water management

The farming is one of my pleasures. I will go to the fields every day, even if farming work is outsourced. I think better yield can be obtained if I grow rice by myself.

I am afraid of possible yield decrease by outsourcing. I want to grow my rice by my way.

Concrete work which can be possibly outsourced

Fertilizing Check of water leakage

Acceptable cost for new service

No additional cost is acceptable No additional cost is acceptable


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.

http://www.sciencedirect.com/science/article/pii/S0378377402001695�



http://www.sciencedirect.com/science/journal/03783774�





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

F-10


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. Materials and Methods 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 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.

Fig. 1 Location of research facilities


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:

)(1 iiiiiii InfETOutRainInWW ++−++= − (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:

HRVINFDRPRi OOOII ++=+ (2)

)PRIINPDRnET IIOOO ++= ( (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.


Table 1 Water balance

3-2. Nutrient balance

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)

Fig. 2 Water balance of paddy field

Time Inflow (mm) Outflow (mm) 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

Fig. 3 Temporal variation of BOD5 and CODCr concentration


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.

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)

Time Input (kg/ha) Output (kg/ha) 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

Fig. 6 Temporal variation of phosphorus concentration


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)

Time Input (kg/ha) Output (kg/ha) 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.

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

Water Quality

Input (kg/ha) Output (kg/ha) Net ouflow loads (kg/ha) Rainfall Irrigation Subtatoal Runoff Infiltration Subtatoal

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


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.





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

F-15


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

i

n

iiiii OOPONSE

1 1

22 )(/)(1 (1)

Where, iO denoted the observed runoff, and iP is the simulated runoff. iO 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

http://en.wikipedia.org/wiki/Calibration�

http://en.wikipedia.org/wiki/Uncertainty_analysis�

http://en.wikipedia.org/wiki/SWAT_model�

http://en.wikipedia.org/wiki/GLUE_(uncertainty_assessment)�


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

GWQMN Threshold depth of water in the shallow aquifer required for return flow to occur Add to value 1.215

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





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;

F-16


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. Materials and methods

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. Results and Discussion 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).

Percent of Days Flow Exceeded (%)

0 20 40 60 80 100

CO

D (k

g/da

y)

0

50100

150

200

Percent of Days Flow Exceeded (%)

0 20 40 60 80 100

TO

C (k

g/da

y)

0

50100

150

200

Percent of Days Flow Exceeded(%)

0 20 40 60 80 100

T-N

(kg/

day)

0

3060

90

120

150

Percent of Days Exceeded(%)

0 20 40 60 80 100

SS(k

g/da

y)

0

200400

600

800

1000

Fig. 5 Observed LDC of COD, TOC, T-N, and SS of the paddy fields.


Persent of Flow Exceeded(%)

0 20 40 60 80 100

BO

D (k

g/da

y)

0

50

100

150

200

Target BOD 2.1 mg/L Target BOD 5.6 mg/L Observed BOD

Persent of Flow Exceeded (%)

0 20 40 60 80 100

T-P

(kg/

day)

0

5

10

15

20

Target T-P 0.151 mg/L Target T-P 0.620 mg/L Observed T-P

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 .

Station

BOD T-P

Target

concentration

(mg/L)

Exceedence

Probability (%)

Target

concentration

(mg/L)

Exceedence

Probability (%)

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





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.

P-03


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. Results and Discussion 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

SMI (cm) ROP (%) BII (cm/d) BFP (%) S-SS (cm/hr) Surface Comp. (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.





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: [email protected]

**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: [email protected]

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.,

P-04


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 of reducing inflow of radioactive substances

Figure 2 Study area in Fukushima, Japan

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. Materials and methods 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

Block 1 2 3 4 5 6 7 8 9 10 Total (ha)

Area (ha) 5.9 25.3 25.2 16.8 11.6 4.8 9.7 60.6 41.4 82.3 283.6

Block 11 12 13 14 15 16 17 18 19 20 Total (ha)

Area (ha) 83.0 26.7 59.0 80.8 18.1 11.8 88.9 15.2 62.6 10.7 456.8


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 upstream

Figure 5 Radiological dosage and turbidity in main canal middle

Table 2. Water management by H land improvement district in the irrigation period of 2013

Month

Discharge May June July August September Total

Control (time) 3 2 5 8 5 23

Stop (day) - - 7 4 7 18

3. Results and discussion

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)





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

P-06


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).


)]12(1[150

−+−= αI

I N

NP (1)

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

IIDD PPP ×= : (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

Parameter unit Agriculture (adults)

Reference Carcinogen Noncarcinogen

AT Average lifetime Days 10950 25550 1)KNSO, 2006

ED Exposure duration years 30 70 2)NIER, 2006

EF Exposure frequency days/year 100 U.S. EPA, 1997

EFf Exposure frequency for food days/year 365 Jang et al., 2007

Bw Body weight Kg 60 2)NIER, 2006

SA Skin surface area cm2 17000 2)NIER, 2006

SAe Exposed skin surface area cm2 5700 U.S. EPA, 2004a

ABS Absorbed fraction from soil to skin unit-less As : 0.03, Cd : 0.001 U.S. EPA, 2004a

AF Soil to skin adherence factor mg/cm2 0.07 U.S. EPA, 2004a

CRw Rate of irrigation water ingestion L/day 0.002 Jung et al., 2005a

CRs Rate of soil ingestion mg/day 100 U.S. EPA, 1997

CRf Rate of food ingestion g/day-capita 215.9 1)KNSO, 2006

FI Fraction of ingestion unit-less 1 -

PC Chemical-specific dermal permeability constant cm/hr 0.00084 U.S. EPA,

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

Irrigation water (paddy water) Irrigation water ingestion Dermal contact in water

Paddy soil Soil ingestion Dermal contact in soil

Air Inhalation


Table 3. Toxic characteristics of target heavy metals

RfD/RfC1) (mg/kg-day) SF2) (kg-day/mg)

5)BCFgrain 6)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. Results and Discussion 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 Coliform 22216.7 58489.2 21713.7 47960.3

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)

Paddy field Irrigation water Ground water WWTP water Reclaimed

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)Concern Level 1)Action Level Paddy rice field 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)KSAW 2)RMC Irrigation water 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

Note) 1) KSAW: Korean Standard for Agricultural Water-quality (the Ministry for Food, Agriculture, Frestry and Fisheries Republic of Korea), 2) 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

U.S. EPA., 1991, Risk assessment guidance for superfund volume I: human health evaluation manual (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.

U.S. EPA., 2004, Risk assessment guidance for superfund volume I: human health evaluation manual (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

Documents

irrigated rice

rural water management

water resources

paddy ecosystem

irrigation practice

water user associations

future water foot

rice quality tasuku