1 YIELD AND WATER USE OF ALTERNATIVE RICE PRODUC T I ON SYSTEMS IN TANZANIA: FIELD EXPERIMENTS AND MODELING B y STANSLAUS TERENGIA MATERU A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLM ENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2014
2 2014 Stanslaus T erengia Materu
3 ACKNOWLEDGMENTS I would like to thank the USAID iAGRI (United State Agency for International Development Innovative Agr iculture Research Initiative) for financial support which enabled me to follow my studies at UF and research part in Tanzania. This support contributed greatly to the successful completion of my Masters studies. I would like to thank my UF supervisor Prof. Sanjay Shukla at the Agricultural and Biological Engineering department and Prof. Andrew K. P. R. Tarimo at the Department of Agricultural Engineering and Land Planning (DAELP), Sokoine University of Agriculture (SUA) for their support and advice which fa cilitated the completion of this work. I would also like to thank my committee members Prof. Water Bowen and Prof. Asseng Senthold for their guidance. I am grateful and appreciate the contribution from Prof. Valerian C. K. Silayo, Prof. Siza Tumbo, Prof. D idas N. Kimaro and the staff at DAELP, SUA. I also appreciate my employer (SUA Management) for providing study leave to pursue my MS degree at UF and providing field site to conduct for my research. I have enjoyed calm working place at DAELP. Special thank s to Naza, Epignosis, Wilbard, August, Anael, Fatina, Nikas, Amani, Mambando, Kihwele and Gasper for the hope they provide to me all the time.
4 T ABLE OF CONTENT S ACKNOWLEDGMENTS ................................ ................................ ................................ .. 3 LIST OF TABLES ................................ ................................ ................................ ............ 6 LIST OF FIGURES ................................ ................................ ................................ .......... 8 LIST OF ABBREVIATIONS/SYMBOLS ................................ ................................ ....... .. 10 ABSTRACT ........................................................................................... 1 1 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ....... 13 Background ................................ ................................ ................................ ........ 13 Justification ................................ ................................ ................................ ......... 14 Goal and Objectives ................................ ................................ ........................... 16 2 WATER USE AND YIELD EVALUATION OF IRRIGATION MANAGEMENT ALTERNATIVES FOR RICE PRODUCT ION IN TANZANIA ................................ ..... 17 .......... 17 Methods and Ma terials 20 Study Area ................................ ................................ ................................ ...... 20 Location ................................ ................................ ................................ ....... 20 Soils and Topographic ................................ ................................ ................. 20 Experimental Design ................................ ................................ ...................... 21 Measurements ................................ ................................ ................................ 23 Climate ................................ ................................ ................................ ......... 23 Soil Physical and Hydraulic Properties ................................ ........................ 24 Plant ................................ ................................ ................................ ............. 24 Soil Moisture and Irrigation M easurements ................................ .................. 25 Data Analysis ................................ ................................ ................................ .. 25 Results and Discussion ........... 26 Plant Gr owth and Yield ................................ ................................ ................... 26 Plant Height ................................ ................................ ................................ 26 Tillers ................................ ................................ ................................ ........... 27 Biomass ................................ ................................ ................................ ....... 27 Yield ................................ ................................ ................................ ............. 28 Crop Water U se ................................ ................................ .............................. 30 Soil Moisture and Irrigation ................................ ................................ .......... 30 Crop Water Use Efficiency ................................ ................................ ......... .. 32
5 Conclusions .. .. ...... 33 3 S IMULATING THE YIELD AND WATER SAV INGS EFFECTS OF RICE IRRIGATION MANAGEMENT ALTERNATIVES IN TANZANIA ................................ ..................... 54 Introducti 54 Materials and M ...... 57 Field Experiment ................................ ................................ ............................. 57 Model Setup, Description and P arameterization ................................ ............. 58 Model S etup ................................ ................................ ................................ 58 Model Description ................................ ................................ ........................ 59 Crop Growth .. 59 Soil Water D ynamics .. ...... 60 Parameterization ................................ ................................ .......................... 61 Mod el E valuation ................................ ................................ ......................... 62 Model Calibration and V alidation .. . 62 Root Mean Square Er ror ... 63 Nash Sutcliffe E fficiency (NSE) ... .. 64 Percent B ias (PBIAS) ... ... 65 Water Management Scenarios ................................ ................................ .... 65 Results and Discussion 66 Model Performance ................................ ................................ ........................ 66 Calibration ................................ ................................ ................................ .... 66 Validation ................................ ................................ ................................ ........ 67 Evaluation of Water Management Scenarios ................................ ................. 68 Conclusions 69 4 CONCLUSIONS AND RECOMMENDATIONS ................................ .......................... 85 8 5 Field Experiment 85 Modeling 8 5 Recommendations 8 7 APPENDI X STATISTICAL PARAMETERS ............ ....................................... .... 8 9 LIST OF REF ERENCE ................................ ................................ ................................ .. 92 BIOGRAPHICAL SKETCH ................................ .............................. 100
6 L IST OF TABLES Table Page 2 1 Mean monthly climatic data (average for 1971 2000) and p otential evapotranspiration (ET) for Morogoro, Tanzania ................................ ................ 50 2 2 Physical soil properties ................................ ................................ ....................... 51 2 3 Important dates and crop develo pments stages for the dry and wet seasons .... 52 2 4 Number of plant height and tillers for the dry and wet seasons .......................... 53 3 1 Input da ta files and descriptions for the ORYZA2000 model .............................. 78 3 2 The c alibrated soil parameters. ................................ ................................ .......... 79 3 3 The c alibrated development rate constant (DRC) (C day 1 ) for dry season ....... 81 3 4 Validated leaf area index (LAI) for the wet season ................................ ............. 81 3 5 Model calibration re sults for simulations of crop growth and yield variables for the dry season (October 2012 January 2013). ................................ ......................... 82 3 6 Model validation results for simulations of crop growth and yield variables for the validation period (wet season (October 2012 January 2013). .......................... 83 3 7 Simulated evapotranspiration (ET), drainage, and change in soil water storage along with irrigation and rainfall for t he calibration (dry season 2012) ................ 84 3 8 Simulated evapotranspiration (ET), drainage, and change in soil water storage along with irrigation and rainfall for the validation (wet season 2012) ................. 84 A 1 Statistical results for comparing LAI values for the dry season ........................ 89 A 2 Statistical results for comparing LAI values fo r the wet season ....................... 89 A 3 Statistical test results for total biomass for the wet season ............................... 89 A 4 Statistical test results fo r total biomass for the dry season ................................ 90 A 5 Tukey Kramer grouping for treatments least squares means (Alpha=0.05) with regards to crop yield for the dry season. ................................ ........................... 90 A 6 Tukey Kramer grouping for treatments least squares means (Alpha=0.05) with regards to crop yield for the wet season ................................ ........................... .. 91
7 A 7 Tukey Kramer test res ults for crop yield differences for dry season ................. 91
8 L IST OF FIGURES Figure Page 2 1 Monthly minimum and maximum temperature for the two seasons 36 2 2 Average monthly rainfall for the two s easons 37 2 3 The d ry seaso n leaf area index (LAI) 38 2 4 The w et season leaf area index (LAI) 39 2 5 Irrigated wate r (a) and 41 2 6 The dry season total biomass.. 42 2 7 The w et 43 2 8 The d ry and wet season rice crop yields for continuous flooding (CF), system of rice intensif ication (SRI), 80% SRI, and 50% SRI.. .. 44 2 9 Daily soil moisture at 30 45 2 10 Daily soil moisture at 30 cm depths durin g the dry season. 46 2 11 The w et season daily average soil moisture below the root zone at 60 cm depth from the s oil surface against dates after t ransplanting . 47 2 12 The d ry season average daily soil moisture below the root zone at 60 cm from the soil surface against dates after transplanting .... 48 2 13 Water use efficiency (WUE).. .. ... .. 49 3 1 Simulated against measured grain yield for the four irrigation treatments for the calibra .. 1 3 2 Simulated against measures total shoot biomass for the four irrigation treatments ( wet season, March 2013 to June 2013) .. 2 3 3 Simulated and measured biomass for different plant organs fo r the four irrigation treatments against days after transplanting (DAT) .. 73 3 4 Simulated and measured LAI of four different water treatments against days after transplanting (DA T) 74
9 3 5 Simulated against measured soil water content at 20cm depth from soil surface 3 6 Simulated and measured ponding depth for different days after transplant (DAT) for the 76 3 7 Simulated crop yield for different irrigation w ater regimes/scenarios 77
10 L I ST OF ABBREVIATIONS/SYMBOLS AWD Alternate Wet and Dry C Celsius Degrees CF Continuous Flooding CRBD Complete Randomized Block Design DAT Day After Transplant DOY Day of the year DSS Decision Support System ET Evapotranspiration FAO Food and Agriculture Organization of the United Nations FSE Fortran Simulation Environment HCl Hydrochloric acid Kg/ha Kilogram per hect are LAI Leaf Area Index mm M illimeters NSE Nash Sutcliffe Effi ciency PBIAS Percent bias RMSE Root Mean Square Error SAS Statistical Analysis System SD Standard Deviation SRI System of Rice Intensification SUA Sokoine University of Agriculture USA United States of America UT Utah WUE Water Use Efficiency
11 Ab stract of Thesis Presented to the Graduate School Of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science YIELD AND WATER USE OF ALTERNATIVE RICE PRODUCTION SYSTEMS IN TANZANIA: FIELD EXPERIMENTS AND MOD ELING By Stanslaus Terengia Materu May 2014 Chair: Sanjay Shukla Major: Agricul tural and Biological Engineering Rice production is important for food security but given its large water footprint, alternative irrigation management strategies are needed. Three irrigation management alternatives were evaluated against the traditional continuously flooded rice (CF) system through field experiment and modeling approaches in Tanzania. The alternative irrigation strategies were implemented with the System of R ice intensification (SRI) practices. The three SRI treatments included applying full, 80% and 50% of the irrigation volume needed for traditional SRI. The experimental evaluation of treatments was conducted in the wet and dry seasons, using a complete rand omized block design. Weather, crop, and soil measurements were made for both seasons. For the dry season, SRI and 80% SRI produced higher yield s of 9.68 tons/ha and 11.45 tons/ha and saved 26% and 35% of water, respectively compared to CF (8.69 tons/ha). T he yield advantage of 80%SRI and SRI over CF for the wet season was less than the dry seasons; 80% SRI and SRI produced 6.01 tons/ha and 5.99 tons/ha and saved 33% and 18% of water, respectively compared to CF (5.64 tons/ha). The 50% SRI had lowest yield o f all, 7.48 tons/ha and 4.99 tons/ha with 54% (693 mm) and 57% (425 mm) water
12 saving compared to CF for the dry and wet season s, respectively. On average, 80% SRI treatment outperformed all other treatments with an additional yield of 1.57 tons/ha and 33.9 % water savings (344 mm) compared to the CF. Field verified rice model (ORYZA2000) was used to evaluate different levels of irrigation input with SRI practices Simulation results for the calibration (dry season) and the validation (wet season) indicated that model performance was satisfactory. The Nash Sutcliffe efficiency (E) for yield predictions ranged from 0.69 to 0.89 for the dry season and 0.60 to 0.83 for the wet season. The E values for simulated soil moisture ranged from 0.68 to 0.89 for the dry season and 0.62 to 0.73 for the wet season. Simulation of irrigation scenarios indicated that irrigating at 75% SRI and 70% SRI produced highest yield for the dry and wet seasons, respectively. Extrapolation of results from this study to Tanzania showed w ater savings of 422,223 ha cm with additional 4.9 million tons of yield
13 C HAPTER 1 INTRODUCTION Background Agricultural water use accounts for 70% of global freshwater water withdrawals (Alvaro et al., 2010). To meet the future increase in deman d for food and fiber production, sustainable water use practices need to be developed and implemented Given the worldwide impacts of irrigation withdrawals, the agricultural community is faced with changing the operational style(s) to conserve water and i ncrease water productivity (Rejesus et al., 2011). Water scarcity is impacting the ability of countries to meet the increased food demand (Zwart et al., 2004). Of the three highest food crops namely, maize, wheat and rice, rice is the most important stapl e food crop in the world (Dawe, 2002). In Asia where 90% of world rice is produced, agriculture water use accounts for 80% of total water use and half of it is used for rice production (Parsi et al., 2003; Behrouz et al., 2010). I t takes 3000 to 5000 L of water for every 1kg of rice produced (Bhuiyani, 1992; IRRI, 1993). Given such a large water footprint of rice practices that can reduce the water use (irrigation volume) of rice and yet sustain or improve the yield need to be developed and tested Similar to other continents, rice production is also impacting water resources in Africa. Africa harvested 9.2 million hectares of rice in 2006; it represents 6.0% of the total rice production area in the world East Africa accounts for almost 25% of the rice pro duction area in Africa ( Chapagain et al., 200 6 ) Increased rice acreage in Africa is impacting other water users such as human, hydropower, and environment. For example, water resources conflict has been experienced in Usangu Plains Tanzania b etween rice f armers (upstream) and Kihansi/Kidatu hydropower stations (downstream)
14 ( Nelson, 2009; Maganga, 200 3 ; Kitova, 2001). Water resources play a key role in as well as provision of ecosystems services for wildlife, tourism and affects livelihood (Noel, 2011). Therefore, there is a need to develop strategies for water conservation in Tanzania The system of rice intensification ( SRI ) has achieved substantial success in producing rice with less water input in several countries including Madagascar, Kenya, Mali, China, and India ( Mati et al., 2011 ). For example t he SRI pilot projects in India have shown substantial increases in crop yields and farm income with 30% less water application (WWF, 2010). In order to contribute towards effort s for sustainable water use and management strategies in Tanzania as well as addressing the expanding land area under rice production due to the ever increasing food demand, there is a need to evaluate potential advantages of water saving practices such as SRI against the traditional continuous flooding (CF) rice production systems. Justification The agricultural water demand in Tanzania exceeds the available water supplies (Katambala et al., 2013). Likewise the demand for food to feed the growing populatio n is increasing which calls for technologies and farming practices to ensure food securities at the same time reducing agricultural water use (Katambala et al., 2013) Rice in Tanzania is produced at subsistence level and most farmers practice continuous f looding a practice that uses large amount of water. S ystem of R ice I ntensification (SRI) a new practice of growing paddy rice which is gaining popularity has been shown to be effective in water saving, increasing yield and has been adopted in many cou ntries (Katambala et al., 2013).
15 The SRI was introduced in Tanzania three years ago but is not yet practiced widely. This practice has been evaluated in Kenya, India, Mali, Madagascar and Bangladesh and could prove to be a potential solution to food securi ty and water conservation Although the alternate wet and dry (AWD) systems have been evaluated, there exist several uncertainties with regards to the level of irrigation needed for optimum yield. Low water input production system evaluated in one region i s specific to the local variety, climate, soil, and water availability and may not be applicable to other regions. There is a need to evaluate strategies that will result in improved yield with lower than current level of water use for CF practiced in Tanz ania. The AWD system has been shown to maintain or increase rice yields in Mali, Kenya, and Madagascar (Mati et al., 2011). However, the AWD is not widely practiced in Tanzania due to lack of scientific evaluation. Farmers in Tanzania need guidelines rega rding the implementation of the AWD system. Without such guidelines and site specific evaluation, farmers and government agencies supporting rice production are hesitant to adopt AWD over the traditional CF system. There is a need to not only evaluate SRI as practiced elsewhere but also evaluate other AWD systems to formulate an optimum irrigation system for rice crop production under Tanzanian conditions. E xperiment al evaluation of AWD water management system is needed to identify the most promising system to be use d under Tanzania conditions. Given the high cost of experimental evaluations in different regions of Tanzania, field verified rice model s using data from experiments can serve as a tool to help evaluate different water management scenarios This will help stakeholders by identifying the optimum production system which addresses the issue of limited water resources. A number of
16 crop growth models which account for soil water dynamics and growth and development of rice are available in literature. T hese models include RICEMOD for potential production and rain fed environments (McMennamy and O_Toole, 1983); WOFOST (Keulen et al., 19 90 ; Hijmans et al., 1994); MACROS model (Penning de Vries et al., 1989) ; and CERES Rice model as part of the DSSAT system (Singh, 1994). Among these models, ORYZA2000 model is a better model because of its capacity to simulate growth and development of rice under water and nitrogen limited environments. Goal and Objectives The overall goal of this study was to establish th e optimum irrigation strategies with regards to yield and water application (irrigation volume) through measurements and modeling. Specific objectives include: (i) conduct a field experiment to evaluate traditional flooded rice against the AWD rice produc tion system with regards to water application and yield and identify the optimum irrigation s trategy for rice production in the dry and wet seasons in Tanzania ; and ( ii ) use the experimental data to field verify ORYZA2000 for identification of the i rrigat ion strategy that sustains or improves the current yield with reduced water use for the dry and wet seasons. Objectives (i) and ( ii) are discussed in Chapter s 2 and 3, respectively. The experiment was conducted during the dry season ( October 2012 to Februa ry 2013 ) and the wet season ( March 2013 to August 2013 )
17 C HAPTER 2 WATER USE AND YIELD EVALUATION OF IRRIGATION MANAGEMENT ALTERNATIVES FOR RICE PRODUCTION IN TANZANIA Introduction Water is a valuable resource and is becoming increasingly scarce (Fran k, 2006). Agriculture is faced with changing practices to conserve water and increase water efficiency. Water scarcity is challenging the ability of countries to meet the increased food demand (Zwart et al., 2004). To meet the future increase in demand for food and fiber production, sustainable irrigation practices need to be developed and implemented. Of the three main food crops, maize, wheat, and rice, rice is the m ost important staple food crop (Dawe, 2002). Given its large water footprint, practices that can reduce water inputs to rice production need to be explored. Deficit irrigation is a technique used to minimize water losses and increase water efficiency espe cially in areas where there is insufficient water supply for irrigation. Deficit irrigation management involves inducing marginal stress, except in critical growth stages where crop yield might be effected (Boggess and Ritchie, 1988). Tanzania has the seco nd largest area (945,087 km) in East Africa and accounts for 6% of the total African rice production area (URT, 1995a). In Tanzania, renewable water supply is estimated to be around 2,300 m3 per person per year (Noel, 2011). Tanzania currently is not clas sified as a nation suffering from water scarcity; however the country is expected to have water scarcity problems by 2015 due to population growth (Noel, 2011). Given this projection, there is a need to develop alternative farming systems that can produce more crops with less amount of water.
18 Rice uses two to three times more water than other cereal crops ( Tabbal et al. (2002 ). In rice production systems, a large quantity of water is lost through evaporation, surface runoff, seepage, and deep percolation ( Guerra et al., 1998). A number of water saving irrigation techniques have been developed for rice (Bouman, 2001). For instance, in Asia, the most widely adopted water saving practice is an aerobic rice production system utilizing raised beds. However, this practice has some limitation related to soil type, rice variety, and socio economic constraints (Mao and Zhi, 1993). Other strategies being pursued to reduce rice water requirements include practices such as saturated soil culture (Borrel et al., 1997). T abbal et al. (2002) found that the combination of system of rice intensification (SRI) agronomic practices and alternating wet and dry (AWD) conditions resulted in higher yield than the continuously flooded system (CF) (Shan et al., 2002). The SRI is a re latively new rice production practice that has been evaluated in countries such as Mali, Madagascar and Kenya in sub Saharan Africa (Mati et al., 2011). The SRI practice is a combination of agronomic practices comprising of land preparation, seed selectio n techniques, nursery establishment, transplan ting of young age seedling of 8 to 12 days, spacing of 25 by 25 cm between rows and plants, alternate flooding to induce wetting and drying (AWD), and frequent weeding (Mati et al., 2011). The SRI practices hav e been reported to have high water use efficiency (Stoop et al., 2002). The AWD is an irrigation practice where water is applied to the field to attain certain depth of ponding after which the field is left unirrigated for some time (e.g. 5 to 7 days) to l et the soil dry out and develop cracks. On the other hand, for CF, water is applied at a frequency that will maintain a certain depth of ponding throughout the season. Under the CF rice production system, more than
19 50% of irrigation water is lost through s eepage deep percolation and excessive unproductive evaporation (Wopereis et al., 199 6 ). The AWD irrigation technique used in the SRI production system has achieved 25 85% reduction in water loss by adjusting the irrigation volume to meet the rice water re quirement and minimiz ing unproductive evaporation and deep percolation losses (Hafeez et al., 2007). Other water saving techniques developed for rice production are: i) selection of heavy soils or areas with hard pan in lower horizons to control losses to groundwater by growing on relatively large rice plots; ii) proper land leveling of rice fields to allow more uniform water distribution; and iii) intensive puddling to reduce soil permeability (Li, 2001). The current SRI practice lacks specific informatio n regarding the irrigation management needed to achieve optimum yield. The SRI combined with AWD has a potential to reduce water application and yet increase or sustain the current yield in Tanzania. However, limited research has been conducted on the eva luation of SRI in combination with different levels of irrigation management in Tanzania. The goal of this study is to address the following questions: i) can SRI increase crop yield and reduce water application compared to CF?; ii) if yes, what are the yi eld gains and water savings ?; iii) is there an irrigation management system that uses less water and yet provides higher yield than CF as well as SRI?; and iv) is it possible to have a higher rice yield with half of the water that is used for the SRI ? The objective of this study was : i) to conduct a field experiment to evaluate traditional flooded rice against the AWD rice production system with regards to water application and yield and ii) use a field verified rice model to identify the optimum irrigati on system for rice production in the dry and wet seasons of Tanzania.
20 Methods and Materials Study Area Location The study was conducted near Morogoro, located 200 km south west of the commercial city of Dar es Salaam. The experimental fields were located at the research farm of Sokoine university of agriculture ( SUA ) located 3.0 km from the Morogoro Municipal area at a latitude of a longitude of of 510 m above mean sea level. The average annual rainfall in Morogoro ranges from 800 mm to 1300 mm and the average annual temperature (Fig ure 2 1) and humidity is 29 C and 79%, respectively (Magan ga, 2003). The dry and wet seasons in Tanzania have different water needs due to differences in rainfall and evaporative demands. Given the weather variability, the optimum irrigation strategy for the dry season is likely to be different from the wet seaso n. This necessitated the evaluation of different irrigation water management practices in both seasons in this study. Although rice can be grown in both the dry (September to January) and the wet (February to June) seasons in Tanzania, limited water availa bility and vagaries in weather and rainfall (Table 2 1) limits the lager scale production of rice during the dry season. Soils and Topographic The research site can be predominantly characterized by dark brown, clay loam, top soils (0 30 cm) with 47% clay, 7% silt, and 46% sand. The volumetric water content at field capacity and permanent wilting point in the 0 30 cm depth is 40.1% and 28.7%, respectively (Table 2 2). The soil is acidic with a pH of 5.6. Other characteristic include: coarse, strong, gr anular, structure; slightly hard when dry, very friable when moist,
21 slightly sticky and plastic when wet, and low CaC O 3 The site landscape c an be characterized by valleys and gentle slopes. Experimental Design The experiment was conducted during the dry a nd wet seasons of 2012 2013. The layout of the experiment was a complete randomized block design (CRBD) with factorial arrangement of four treatments with three replications. A total of 12 experimental plots, with an area of 40 m 2 each were located at the experimental site. The study evaluated four production systems (treatments) which varied in irrigation volume between transplant and before panicle initiation stage (Table 2 3). The different treatments included 1) continuous flooding irrigation regime (C F); 2) the SRI (maintain 40 mm ponding for three consecutive days followed by no irrigation for five days); 3) the 80% of irrigation volume of SRI (32 mm ponding for three consecutive days followed by no irrigation for five days); and 50% of irrigation vol ume of SRI (20 mm ponding for three consecutive days followed by no irrigation for five days). In the CF treatment, irrigation water was applied to maintain a 50 mm ponding depth throughout the two seasons. All SRI systems involved alternate wetting and d rying (AWD) during the initial st ages (Table 2 3) After panicle initiation stage, continuous flooding was practiced in all treatments to maintain ponding until plant senescence to ensure grain filling. For this study, 80% SRI treatment is a 20% reduction in volume of irrigation applied on SRI Seeds for the experiment were obtained from the Africa seed center at SUA, Morogoro. The seeds were placed in a bucket containing salt solution and the ones which settled were used for the experiment as suggested by Mati et al. ( 2011). SRI
22 treatments involved transplanting 12 day old seedlings, with two leaves. The seedlings were trans planted carefully and quickly to minimize roots stress. Thereafter, the field was left on a mud form for four days to allow establishm ent. The number of seedling per hill was one for all SRI treatments; this allowed optimum growth without competition for nutrients (Sakurai. 2006). Typically farmers in Tanzania transplant three seedlings per hill in CF system (Mati et al., 2011). To encou rage greater root and canopy growth, plant to plant and row to row spacing was maintained at 25 cm for all SRI treatments (Dobermann, 2004). Standard steps for the SRI practice (Mati et al., 2011) followed in this study are described below. Field prepar ations : T he soil was kept saturated for 5 days, and then rotavated. The field was harrowed twice at an interval of 3 days to ensure proper soil water mixture. Fertilizer was applied before the last puddling event. Each plot was leveled to allow uniform wat er ponding over the plot. Nursery : A 3 cm thick seedbed made of a mixture of soil and organic fertilizer (1:1) was placed on top of a plastic sheet. Rice seeds were sown in the seedbed The sheet prevents seedling roots from getting into the soil. The sow ing rate was 7 kg/ha. Irrigation was done two days after sowing to keep the soil saturated, but not flooded. Transplanting : Twelve day old seedlings were transplanted before the emergence of a third leaf. Care was taken to separate seedling from the seedb ed to avoid damage to the young root. Seedlings were quickly transported on a tray to the field to avoid stress. One seedling was planted per hill at a depth of 2 cm on the grid (25 x 25 cm). Irrigation : Between the transplanting and appearance of panicle s, three to five days irrigation cycle was followed, i.e. the field was irrigated for three consecutive days
23 and then left to dry for five days. The goal was to keep the soil moist but not saturated to allow air to get into the soil for improved soil healt h and root growth. After panicle initiation, irri gation was applied to maintain 2 cm ponding for all three SRI treatments. Weeding and fertilizer : Weeding was done every 12 days. A spike toothed rotary tool was used for manual weeding and to aerate the so il. Fertilizer was applied to meet the plant nutrient requirements for optimum yield at the rate of 500 kg/ha (N), 50 kg/ha (K 2 O) and 50 kg/ha (P 2 O 5 ). Harvesting: A t grain maturity, the field was drained to allow the soil to dry before harvesting. Measurem ents Climate Annual rainfall was measured at the site using a standard rain ga u ge. Another set of rainfall measurements were taken at the meteorological station at SUA (Fig ure 2 2) Humidity, wind speed, solar radiation, and temperature data were obtained from the SUA meteorological station and were used to calculate reference evapotranspiration (ET O ) using FAO Penman Monteith equation (Allen et al., 1998) (Table 2 1). The average annual temperature at the site is 23 o C with a minimum of 15 o C in July and a maximum of 32 o C in November and December (Fig ure 2 1. The seasonal mean relative humidity was 66 % and 78 % for the dry and wet seasons, respectively. The mean relative humidity ( 1971 2000 ) for the area is 73% (Table 2 1) (FAO. 1998 ). Rainfall is bimodal, characterized by two rainfall peaks with short rains from October to December and long rains from March to May. Total rainfall during the dry season was 489 mm and 1419 mm during the wet season (Fig. 2 2 ).
24 Soil Physical and Hydraulic Properties Percentag es of sand, clay and silty were determined by using the sieving method from the soil sample taken from the site and other physical properties such as f ield capacity, permanent wilting point, saturated hydraulic conductivity were calculated (Table 2 2). Pla nt Five plants were randomly selected from each plot for measuring height and number of tiller s The yield from each plot was weighed at the end of each season. Three, 1 m quadrants were selected in each plot for yield measurements. Dry biomass (oven drie d for 24 hours or more until no change in weight) of different plant organs (stem, leaves and panicles) were weighed. Length and width of leaves were measured manually to estimate leaf area index (LAI) as suggested by Xinyou, et al. (2000 ) ( Table A 1), (Ta ble A 2), (Fig ure 2 3) and (Fig ure 2 4) Equation (1 1) e xpresses the LAI as calculated by Xinyou. (1 1) Where; LAI is a leaf area index, Wlvg is the green leaf weight and SLA is the specific leaf area T he SLA was calculated as: (1 2) Where; a, b, c and d are the functional numeric parameters for rice (a= 0.0024, b= 0.0025, c= 4.5, and d= 0.14 ) a nd DVS is the development stage in days.
25 Soil moisture and Irrigation measurements Irrigation water volume for each treatment was measured using a propeller type flow meter. A flow control valve was i nstalled at the main water line of the farm to control irrigation volume for each plot. Volume of each irrigation event was measured. A polyvinylchloride pipe was connected to the flow meter with an attached flexible hose pipe to deliver water to individua l plots. Deep percolation for each plot was estimated using ORYZA2000 model with following assumptions: i) rainfall (Fig ure 2 2 ) and irrigation (Figure 2 5 a) were the only source of water; ii) soil reservoir responds to water application by reaching equili brium instantaneously meaning that, after equilibrium, the excess water percolates to recharge the ground water; and iii) the contribution of groundwater to the root zone through capillary rise is negligible. The water table at the site was 2 m below the gr ound surface Soil moisture was measured on a 15 minute interval basis using one capacitance probe (sensors at 10 cm, 20 cm, 30 cm, 60 cm, 80 cm, and 90 cm below the soil surface) in one of the plot s for each of the four treatments. The probe s were connect ed to a CR206 datalogger (Campbell Scientific Inc., Logan, UT, USA) to store the 15 min ute data from which data were downloaded to a laptop. For the 2012 2013 dry growing season soil moisture was measured from November 3, 2012 to January 30 2013. During t he wet season, soil moisture data could only be measured from March 4 May 6, 2013 due to theft of the datalogger. Data Analysis A ll statistical analyses were conducted using SAS (SAS, 1990). The data were analyzed using Tukey Kramer test. Tukey Kramer is a technique of multiple comparison procedure of pairwise differences of mean parameters (Somerville, 1993). Variables
26 compared included water applied (mm), crop yield (kg/ha), soil moisture (%vol.), LAI, above ground biomass (kg/ha) and water use efficien cy (WUE, kg/m 3 ). The WUE was calculated using the crop yield (Y, kg) and ET c ( Mostafazadeh Fard 2010 ) as: (2 3) The ET c was calculated as: (2 4) w here ; ET o is the reference evapotranspiration and K c is crop coefficient for rice K c for rice was obtained from FAO ( 1998 ). Results and Discussion Plant Growth and Yield Plant Height For both the dry and wet seasons the plant height for CF and SRI treatments were significantly (p< 0.05) higher than 80% SRI and 50% SRI (Table 2 4). For both seasons these observations can be attributed to higher water uptake due to water availability almost throughout the season The shorter plant height for 80% SRI and 50% SRI implies that more water ponding at the vegetative stage affect s the spike characteristics (Panda et al., 1997; Anbomozhi et al., 1998). When water ponding at this stage is low, it increases the tillering ability of the plant. Similar plant hei ghts for CF and SRI are likely due to similar soil moisture in the root zone. Plant height for the dry season for CF was 18% and 39% higher than 80% SRI and 50% SRI, respectively. In the dry season, SRI was 32 % and 10% greater than 50% SRI
27 and 80% SRI, re spectively (Table 2 4). Plant height in the wet season for CF was 11% and 30% greater than 80% SRI and 50% SRI, respectively. The reason for higher plant height for CF is that with more ponded water, plant nutrient uptake increases which increases plant he ight (Ch a udhary 2003 and Parihar 2004). Tillers For number of tillers, 80% SRI were significantly higher (p < 0.05) than rest of the three treatments for both seasons (Table 2 4). Shortening the vegetative stage duration has been shown to result in inc reased tillers (Panda et al., 1997). Based on the statistical analyses results, the number of tillers can be arranged as 80% SRI > SRI= CF > 50% SRI There was no difference in tillers for CF and SRI treatment (Table 2 4). For the dry season, 80% SRI had 4 0% more tillers than CF and 93% more tillers that 50% SRI. Number of tillers per hill for 80% SRI indicates higher potential yield than the rest of treatments. Because panicles are attached to tillers, number of tillers are usually an indicator of yield; h igher the number of tillers, higher the potential for increased yield. The advantage of the SRI method in enhancing tiller numbers has been observed by Gaini et al. (2002). Results for the wet season were similar to the dry season however the numerical di fferences between the treatments were much lower. Part of this difference was due to rainfall which had the effect of differential irrigation input. Overall, 80% SRI outperformed all other treatments in number of tillers indicating higher yield potential t han other treatments. Biomass Trends in biomass (Fig ure 2 6 and 2 7; Table A 3 and A 4), another indicator of crop yield, mimic ked the trends for LAI (Fig ure 2 3 ), (Fig ure 2 4) ( Table A 1) and (Table
28 A 2) and tillers (Table 2 4). Higher biomass leads to h igher accumulation of non structural carbohydrate in the culms and leaf cover which can rapidly be trans located to the panicle during the initial stage of grain filling which can increase the potential for higher crop yield (Toshiyuki et al., 2006). Based on the results from the statistical analyses, the order for the wet and dry season s biomass were 80% SRI> SRI> CF> 50% SRI (Fig ure 2 7; Table A 4) and 80% SRI> SRI> CF> 50% SRI (Fig ure 2 6, Table A 5), respectively. Overall biomass as well as all other gr owth indicators indicates best plant performance for the 80% SRI treatment followed by SRI for both seasons. Plant growth indicators indicate equal or better performance of the SRI and 80% SRI treatments compared to the CF. Yield For both seasons, the 80% SRI had statistically higher yield than other treatments while the 50% SRI had the lowest yield (Fig ure 2 8, Table A 5 and A 6). For both seasons following comparisons were significant: 80% SRI > CF, CF > 50% SRI, SRI > 50% SRI. For the dry season yield for 80% SRI was significantly higher than the conventional SRI. However, frequent rainfall events between transplanting and panicle ini tiation during the wet season March May 2013 ( Fig ure 2 2) resulted in similar soil moisture in the root zone for all SR I treatments which resulted in similar yields for SRI, 50% SRI, and 80% SRI (Table A 6) and (Table A 7) The w et season yields were lower than dry season mainly due to near saturation to saturated soil moisture for all treatments (Fig ure 2 9) because of frequent rainfall between March and May 2013 (Fig ure 2 2). Although soil moisture values shown in Figure 2 9 show the relative differences in soil moisture, these values may not be
29 accurate representation of actual soil moisture due to sensor errors and/or inadequate installation which may have resulted in bypass vertical flow along the tube. The smallest yield observed for the 50% SRI for the dry season was higher than highest yield for 80% SRI during the wet season. This difference in yield between the dr y and wet season s was due to very little rainfall during November December, 2012 (Fig ure 2 2) which helped maintain the target soil moisture for all three SRI treatments (Fig ure 2 9). In contrast, frequent rainfall during March to May 2013 resulted in si milar soil moisture for all SRI treatments leading to similar yields. The 80% SRI treatment produced 32 % and 7% more rice than CF for the wet and dry seasons, respectively. Although the 80% SRI treatment had almost same yield as SRI for the wet season, it had 18% more yield than the SRI treatment for the dry season. Higher yield for 80% SRI is in agreement with observations from other studies (Krishna et al., 2008 ; Vijayakumar et al., 2001 and Ga i ni et al., 2002) which noted higher grain yield when younger s eedlings (8 to 12 days old) are transplanted at spacing ranging from 25 cm x 25 cm to 30 cm x 30 cm under non flooded conditions. In this study, younger seedling, transplant spacing, and AWD irrigation were the main factors that increased the tillering abi lity (per hill and per area), panicle length, number of filled grains, and finally high yield for the 80% SRI treatment followed by SRI compared to CF. Rice yield for the wet season was low with small differences among the treatments due to heavy and fr equent rainfall at the beginning of the wet season (March and May) (Fig. 2 8) which resulted in sustained saturation/flooding during the wet period and prevented implementation of SRI treatments. Similar results were observed by Stoop at
30 el. (2002), who no ted that it was not possible to attain higher yields with SRI compared to CF due to frequent rainfall events. High plant density (3 to 4 seedlings per hill) and small plant to plant and row to row spacing leads to nutrient competition combined with limited soil aeration, and stress to younger roots during transplanting and contributes to lower crop yield under CF. Overall, 80% SRI consistently outperform ed all treatments in yield across seasons. For the dry season, higher yield for 80% SRI and SRI were due to: 1) adequate soil moisture required by the plant between transplanting and panicle initiation stages which enhances nutrients uptake; 2) reduced plant stress due to non waterlogged conditions in the root zone which promotes healthier root growth; and 3 ) improved soil aeration which increases microbial metabolism activity. These factors resulted in better tillering during the vegetative stage observed in this study (Tables 2 4). Crop Water use Soil Moisture and Irrigation Measured soil moisture within a nd below root zone (60 cm) indicates the success of treatment implementation. During the dry season, soil moisture of all SRI treatments fluctuated around soil field capacity from the transplanting to the panicle initiation stage after which soil moisture for all treatments was similar due to sustained flooding (Fig ure 2 10). Unlike the dry season (Fig ure 2 10); soil in all SRI treatments was near saturation point due to frequent rainfall events (Fig ure 2 2) and (Fig ure 2 11). During the dry season, the onl y treatment which was allowed to fall below field capacity and, at times, close to wilting point was 50% SRI. The soil moisture below root zone at 60 cm for both
31 dry and wet seasons was near or at saturation for 80% SRI and SRI treatment from panicle initi ation to crop senescence stage (Fig ure 2 11) and (Fig ure 2 12). After panicle initiation stage, the soil moisture at 60 cm depth for all SRI treatments for the dry season was similar to CF because of continuous ponding maintained for all the treatments G i ven that 80% SRI resulted in maximum yield, it can be inferred that the desired soil moisture from transplant ing to panicle initiation stage within the root zone is between 44% 48% (vol ) assuming that soil moisture within the 0 30cm represents the soil moisture within the root zone Unlike the dry season, for the entire duration of the wet season soil below root zone was saturated for all the SRI treatments (Fig ure 2 11) because of the heavy and frequent rainfall (Fig ure 2 2). T he desired SRI irrigation cycle (wetting for 3 consecutive days and drying for 5 days) was interrupted by heavy tropical rainfall (at 30 DAT) (Fig ure 2 2) which negated the effects of the irrigation management and led to similar soil moisture in the root zone for all treatments re sulting in reduced yield for SRI treatments. Similar observations were made by Kombe (2012). There was a significant difference in the amount of water applied between the two seasons which led to differences in WUE. Both 50% SRI and 80% SRI treatment s had higher water use efficiency than CF during both seasons. In the dry season, total rainfall was 489 mm. Irrigation volumes applied to the 80% SRI and CF treatments were 830 mm and 1286 mm, respectively (Fig ure 2 5a). On the other hand, during the wet seaso n, there was a total of 1419 mm rainfall and the irrigation volume applied to 80% SRI and CF were 524 mm and 787 mm, respectively (Fig ure 2 5a). The irrigation varied depending on ET and deep percolation losses. Lower irrigation volume for the wet
32 season c ompared to the dry season was due to frequent rainfall (Fig ure 2 5a ) and ( Fig ure 2 2). Results show that using 80% of the irrigation volume of the SRI can save 35% water compared to CF during the dry season ( Fig ure 2 5a). The water savings during the wet season will vary depending on rainfall. On similar lines, Tabbal et al. ( 200 2), reported that maintaining the soil moisture by alternate wetting and drying reduced irrigation volume by about 40 70% compared with the traditional CF without any significant loss in yield. In Tanzania, water abstracted from rivers for irrigation accounts for 56% of the wet season river flows and 93% of the dry season river flows (Mwakalila, 2005). Results from this study indicate d that using 80% SRI for both seasons can achie ve significant water savings which can be used by other users and help maintain environmental flows Crop Water Use Efficiency T he SRI treatments had higher WUE than CF production system indicating higher SRI treatments (Fig ure 2 13). For the dry season, the WUE for the 80% SRI was twice that of the CF (Fig ure 2 13). The 80% SRI treatment was 26% more efficient than SRI during the dry season (Fig ure 2 13). Despite more frequent rainfall during the wet seas on, 80% SRI was still 38% more efficient than CF (Fig ure 2 13). Considering the yield advantage from 80% SRI for both seasons, it is a better irrigation management strategy compared to SRI as well as CF (Fig ure 2 13). Although the WUE for the 50% SRI was h ighest for the wet season, the yield loss from this treatment is not likely to result in acceptance of this treatment over 80% SRI. However, high WUE for 50% SRI during the wet season shows that for areas with lower
33 irrigation supply 50% SRI is still a via ble production system because yield reductions from 50% SRI were only 14% compared to CF. In areas with limited water availability for irrigation, farmers can opt to use 50% SRI management strategy during the wet season as long as rainfall frequency and ti mings are similar to those observed during this study. Conclusions Results show that there is no yield advantage from the traditional continuously flooded (CF) rice production system over the SRI production system for both dry (Oct ober February ) as well as the wet (Feb ruary Aug ust ) growing seasons in southern Tanzania. Consistently, highest yields were obtained from the 80% SRI system for both seasons which indicates that it is possible to increase yields while reducing the total irrigation volume. The 80% SRI system outperformed the CF system by 2762 and 365 kg/ha with water savings of 456 and 263 mm for the dry and wet seasons, respectively. The water savings from 80% SRI is 35 48% compared to the CF system. Given that 50% SRI system produced almost 90% y ield compared to CF during the wet season, the farmers in limited water supply regions are still likely to achieve a viable yield. For the farms that grow rice in both seasons, the annual water savings from the 80% SRI will be 719 mm with an additional pr oduction of 2827 kg/ha over the CF system. Similar results were also observed by Champagain et al. (2011) for the SRI with 20 50% water savings compared to CF. Considering that water savings from the 80% SRI accounts for 77% of annual rainfall (935 mm/yr. ) in southern Tanzania, this irrigation management system has important implications for water supply and maintaining environmental flows in rivers. The water savings realized from 80% SRI is due to reduction in unproductive field losses (evaporation and d eep percolation).
34 The yield advantage and water savings from the 80% SRI is likely to vary depending on rainfall amount and distribution, soil properties, water availability, and management strategies. In 2011, Tanzania cultivated 1,119,324 ha of rice on two season growing basis utilizing mostly the CF production system. Assuming that results from this study are applicable to all rice producing area s of Tanzania, implementation of 80%SRI will result in annual water saving of 402,117 ha cm of water volume a nd achieve additional production of 4.6 million tons of rice. Achieving these water savings and yield benefits is likely to increase the sustainability of rice production system in Tanzania and create additional water supplies for industry, environment, an d other users. In order to realize the yield and water saving benefits from this study in Tanzania for the two seasons, there is need to develop easy to understand water management recommendations for farmers. The AWD irrigation technique for the 80% SRI t ranslates to 45 48 % soil moisture in the root zone during the transplanting panicle initiation period to achieve the yield and water saving benefits. Considering the economic status of most farmers, the use of soil moisture equipment for active management of irrigation may not be a feasible tool. Based on results, applying average irrigation of 395 and 217 mm water for the dry and wet season s before the panicle initiation stage has the potential to achieve increased yield. One of the avenues of promoting la rge scale implementation of the SRI system is to conduct on farm demonstration studies. Furthermore, socio economic factors including market prices, soil type, water availability, and existing irrigation infrastructure will have to be considered for wide s cale acceptance of 80% SRI in Tanzania. Given the projected water shortage by 2030 and large scale production of
35 rice in Tanzania, 80% SRI has potential to improve the well being of farmers and contribute to food security in Tanzania.
36 Figure 2 1 Monthly minimum and maximum temperature for the dry and wet season s 26 25 26 23 25 25 23 24 22 20 23 16 15 17 17 16 18 17 16 13 13 13 0 5 10 15 20 25 30 Temperature ( C) Month Max.Temp. Min. Temp.
37 Figure 2 2 M onthly measured rainfall for the dry and wet season s 0 20 40 60 80 100 120 140 160 Rainfall (mm) Month
38 Figure 2 3 The d ry season leaf area index (LAI) for continuous flooding (CF), system of rice intensification (SRI), 80%SRI, and 50% SRI treatments for different days after transplant (DAT) and crop stages. DAT 0 to 46, 47 to 59, 60 to 72, and 73 to 90 refer to vegetative, panicle initi ation, flowering, and senescence stages, respectively. (Dry season, October 2012 to January 2013). 0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 0 10 20 30 40 50 60 70 80 90 100 LAI DAT CF SRI 80%SRI 50%SRI
39 Figure 2 4 The w et season leaf area index (LAI) for continuous flooding (CF), system of rice intensification (SRI), 80%SRI, and 50% SRI treatments for different days after transplant (DAT) and crop stages. DAT 0 to 46, 47 to 60, 61 to 75, and 77 to 90 refer to vegetative, panicle initiation, flowering, and senescence stages, respectively. (Wet season, February 2013 to June 2013 ) 0 1 2 3 4 5 6 7 8 9 0 10 20 30 40 50 60 70 80 90 LAI DAT CF SRI 80% SRI 50% SRI
40 Figure 2 5. Irrigation water (a) and total water input (b) (irrigation + rainfall) for continuous flood (CF), system of rice intensification (SRI), 80% SRI and 50% SRI treatments for the dry and wet seasons 1286 949 830 593 787 649 524 322 0 200 400 600 800 1000 1200 1400 CF SRI 80% SRI 50% SRI Irrigation (mm) Treatments a Dry season wet season
41 Figure 2 5 C ontinue 1774.7 1438.4 1319.2 1082.1 2166.3 2059.5 1933.3 1941.3 0 500 1000 1500 2000 2500 CF SRI 80% SRI 50% SRI Total water input (mm) Treatments b Dry season wet season
42 Figure 2 6 The d ry season total biomass for continuous flooding (CF), system of rice intensification (SRI), 80% SRI, and 50% SRI treatments for different days after transplant (DAT) and crop stages. DAT 0 to 46, 47 to 60, 61 to 75, and 77 to 94 refer to vegetative, flowering, panicle initiation, and senescence stages, respectively. (Dry season, October 2012 to January 2013) 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000 0 10 20 30 40 50 60 70 80 90 100 Biomass (kg/ha) DAT flood SRI 80%SRI 50%SRI
43 Figure 2 7 The w et season total biomass for conti nuous flooding (CF), system of rice intensification (SRI), 80% SRI, and 50% SRI treatments for different days after transplant (DAT) and crop stages. DAT 0 to 46, 47 to 60, 61 to 75, and 77 to 94 refer to vegetative, flowering, panicle initiation, and sene scence stages, respectively. (Wet season, February 2013 to June 2013) 0 2000 4000 6000 8000 10000 12000 0 10 20 30 40 50 60 70 80 90 100 Biomass (kg/ha) DAT CF SRI 80% SRI 50% SRI
44 Figure 2 8 Dry and wet season s rice crop yields for continuous flooding (CF), system of rice intensification (SRI), 80% SRI, and 50% SRI 8690 9680 11452 7482 5644 5994 6009 4992 0 2000 4000 6000 8000 10000 12000 14000 CF SRI 80% SRI 50% SRI Crop Yield (kg/ha) Treatments Dry season wet season
45 Figure 2 9 Daily soil moisture at 30 cm depth during the wet season CF is continuous flooding SRI is system of rice intensification 80% SRI is 80% of irrigation of SRI and 50% S RI is 50% of irrigation of SRI. DAT is days after tran splant (Wet season, February 2013 to June 2013) 28 33 38 43 48 53 58 0 10 20 30 40 50 60 70 Soil moisture (% vol.) DAT 80% SRI SRI CF 50% SRI
46 Figure 2 10 Daily soil moisture at 30 cm depth during the dry season CF is continuous flooding, SRI is system of rice intensification, 80% SRI is 80% of irrigation of SRI and 50% SRI is 50% of irrigation of SRI. DAT is days after transplant (Dry season, October 2012 to January 2013) 29 34 39 44 49 54 59 64 0 10 20 30 40 50 60 70 Soil moisture (%vol.) DAT ----80% SRI SRI CF 50%SRI
47 Figure 2 11 The w et season daily average soil moisture below the root zone (60 cm) against da ys after transp lanting (DAT) CF is continuous flooding SRI is system of rice intensification 80% SRI is 80% of irrigatio n of SRI and 50% SRI is 50% of irrigation of SRI (Wet season, February 2013 to June 2013) 47 48 49 50 51 52 0 10 20 30 40 50 60 70 Soil moisture (%vol.) DAT CF SRI 80% SRI 50% SRI
48 Figure 2 12 The d ry s eason daily average soil moisture below the root zone (60 cm) against days after transplanting (DAT). CF is continuous flooding, SRI is system of rice intensification, 80% SRI is 80% of irrigation of SRI, and 50% SRI is 50% of irrigation of SRI (Dry seaso n, October 2012 to January 2013) 33 35 37 39 41 43 45 47 49 51 53 55 0 10 20 30 40 50 60 70 80 90 100 110 120 Soil moisture (%vol.) DAT CF SRI 80%SRI 50% SRI
49 Figure 2 13 Water use efficiency (WUE) ; CF is continuous flooding, SRI is system of rice intensification, 80% SRI is 80% of irrigation of SRI, and 50% SRI is 50% of irrigation of SRI treatm ent for the dry and wet season s 169 255 345 315 179 231 287 388 0 50 100 150 200 250 300 350 400 450 CF SRI 80% SRI 50% SRI WUE (kg/m3) Treatments Dry season wet season
50 Table 2 1 Mean m onthly c limatic data (average for 1971 2000) and potential evapotranspiration (ET) f or Morogoro Tanzania Month Max. Temperature (C) Min. Temperature (C) Relative H umidity ( % ) Wind Speed ( km/day ) Sunshine Hours ( h/day ) Solar Radiation ( MJ/m^2/day ) ET o (mm/day) Precipitation (m m) January 31.5 21 .0 71.4 130 5.7 3 18.62 4.23 105 February 31.7 20.8 72.7 121 5.9 0 19.1 0 4.24 97 March 31.5 20.8 76.3 121 5.96 18. 83 4.09 133 April 29.6 20.4 83 .0 103 4. 58 15.64 3.28 198 May 28.2 18.8 82.3 112 4. 2 2 13.92 2.88 79 June 27.3 15.9 78 .0 1 30 4. 46 13.55 2.75 19 July 27.2 15 .0 74.1 130 4. 36 13.67 2.79 13 August 28.3 15.8 69.2 1 30 4.4 1 14.77 3.17 11 September 29.8 16.6 66.9 156 4. 8 9 16.61 3.75 20 October 31.2 18 .0 65.1 190 5. 77 18.63 4.42 43 November 31.8 19.5 67.8 17 3 6.1 4 19.23 4.48 98 December 31.8 21 .0 69.2 173 5.7 4 18.46 4.43 119 Source: Chapagain (2006)
51 Table 2 2 Physi cal soil properties at the experimental field. Horizon Ap Bt1 Bt2 Bt3 Bt4 Bt5 Depth (cm) 0 30 30 55 55 77 77 100 100 130 130 190+ Clay % 47 61 61 67 71 69 Silt % 9 9 11 9 9 7 Sand % 46 30 28 24 20 24 Saturation (% volume) 45.3 48.5 49 .0 49.8 50.6 49.8 F ield capacity (%volume) 40.1 40.7 40.8 40.9 41.1 40.9 Wilting Point (% volume) 28.7 28.3 28.2 28.2 28.1 28.2 Sat urated Hydraulic Conductivity (mm/hr.) 0.35 1.18 1.34 1.68 2.09 1.68
52 Table 2 3 Important date s an d crop developments stages for the dry and wet seasons Event Growth Stage Period (DAT) Dry season date Wet season date Field Preparations Seeds 9/15/12 2/8/13 Nursery seeds 9/24/12 2/18/13 Transplanting seedlings 10/6/12 3/1/13 Tillering V egetative 0 46 11/21/12 4/16/13 Panicle initiation Panicle initiation 47 59 12/3/12 4/29/13 Flowering Flowering 60 72 12/15/12 5/13/13 Grain filling Senescence 73 90 1/1/13 5/27/13 Harvesting Maturity 91 116 1/26/13 6/19/13 DAT is day s after transplant
53 Table 2 4 Number of plant height and tillers for the dry and wet seasons Treatments ** Plant Height (m) Number of Tillers Dry Season Wet Season Dry S eason Wet Season CF 0.49 a 0.52 a 40 a 28 a S RI 0.44 a 0.48 a 42 a 31 a 80% SRI 0.40 b 0.47 b 56 b 38 b 50% SRI 0.30 c 0.40 c 29 c 25 c Treatments with the same superscripts were significant ly different at 0.05 levels ** CF Continuously flooded, SRI System of Rice Intensification, 80% SRI 80% of SRI i rrigation water, and 50% SRI indicates 50% of SRI irrigation water applie d
54 CHAPTE R 3 SIMULATING THE YIELD AND WATER SAVINGS EFFECTS OF RICE IRRIGATION MANAGEMENT ALTERNATIVES IN TANZANIA I ntroduction Overuse of fresh water by agricultural activities has reached to an extent that it is limiting the contribution of other sectors like power production, domestic use and environment to human development worldwide (Janssen and Lennartz, 2007). Inefficiency in irrigation has been the source of large losses in a gricultural productivity (Maganga 2003 ; Kitova, 2001). This loss in productivity combined with need to feed a large world population has led to increase in agricultural area ( Chapagain et al., 200 6 ). There is a need to develop efficient water management s trategies to maintain or increase productivity while reducing the unproductive losses. Half of the agricultural water is used for rice production (Parsi et al., 2003) Among irrigated crops, rice uses most water than all other crops (Maganga 200 3 ). The global average water footprint of rice is 3000L to 5000L per 1 kg of rice (IRRI, 1993). In Africa increase in rice acreage demands the use of more fresh water for its production to feed the increasing populat ion considering that most of it is grown under flood irrigation method (Nelson, 2009). The System of Rice intensification (SRI) imbedded with the Alternate wetting and dry water management has show n higher yield with lower water use compared to traditional flooded rice at several locations such as Ch ina, Madagascar and Kenya (McDonald at el ., 2006). The SRI promote s soil aeration, healthier root system s beneficial microbial activities yield while conserving water A lthough SRI has been promoted there is little information on specific irrigation man agement techniques In
55 chapter 2 among four irrigation management system s evaluated, 80% SRI was found to be the best in terms of yield and water use H owever, it is possible that the optimum yield is either less or more than t h at of 80%SRI. Due to high c ost involved in conduct ing experiments for identifying the optimum water management strategy, modeling is a useful to ol to simulate the effects of a variety of water management strategies for r ice production in Tanzania. For designing an optimum water management strategy, modeling can serve as an effective tool to identif y the sustainable rice production system. A variety of crop growth models have been developed and used throughout the world for improving crop production systems. Crop models can help c ompare findings from experimental research across sites, extend the experimental field data to wider environments, and help to develop decision support systems to evaluate impacts of climate change on yields (Bouman et al., 1996). Although there are a vari ety of crop models, only a few of them can effectively simulate crop growth as well as water and nutrient fluxes. Examples of early rice growth models include RICEMOD (McMennamy and O_Toole, 1983), WOFOST (Keulen et al., 19 90 ; Hijmans et al., 1994) and MAC ROS (Penning de Vries et al., 1989). In the mid 90s, International rice research institute (IRRI) and Wage ningen University developed the ORYZA1 model for simulating growth of tropical lowland rice ( (Kropff et al., 1994a, Ten Berge and Kropff, 1995). The O RYZA1 model was followed by ORYZA_W for water limited production systems (Wopereis et al., 1996) and ORYZA N (Drenth et al., 1994) and ORYZA1N (Aggarwal et al., 1997) for nitrogen limited conditions. Kropff et al. (1994a, 1995) evaluated ORYZA1 by comparin g observed and simulated yields and end of season biomass v alues. However, they
56 conducted limited validation of the dynamic simulation of crop growth variables. The current version of this model, ORYZA2000, is an explanatory and dynamic eco physiological s V an Ittersum et al., 2003). Using results from a field experiment at IRRI, Wopereis (1993) evaluated ORYZA_W and concluded that the model performance was satisfactory. Drenth et al. (1994) ev aluated ORYZA N using seven field experiments spanning different varieties, years, and locatio ns and concluded that ORYZA N model can be used to simulate rice growth for water and nitrogen stress es Aggarwal et al. (1997) evaluated ORYZA 1 N by graphical co mparison of experimental data from three field experiments at IRRI and found that the model was able to simulate the crop growth and development. Bouman et al. (2001) integrated all previous versions of the ORYZA models in ORYZA2000 which was released in 2001. ORYZA2000 use s data for a range of varieties, growth conditions (potential, nitrogen limited, water limited), climatic and geographical locations (Philippines, Indonesia, China). The model assumes that the crop is free from disease, pests and weeds. The ORYZA2000 can simulate rice growth and production in nitrogen limited and water limited environments. The ORYZA2000 model tracks a calculation order for the rate of biomass production and for the rate of phenological development. ORYZA2000 can be combi ned with measurements to simulate and design water management strategy that can optimize yield with least irrigation volume (Bouman et al., 2001). Crop and soil water variables were used to calibrate and validate the ORYZA2000 crop growth model (Bouman et al., 2001), and then use the model to simulate different water regimes The model has been used for simulating rice
57 production under a variety of climatic and production conditions from semi desert s (upland rice) to wet lands ( low land rice or flooded rice ) (Bouman et al., 2001; Van Ittersum et al., 2003). However, limited attempts have been made to evaluate this model in rice producing regions of Africa especially areas such as Tanzania with semi arid tropical climate. The goal of this study was to evaluate the water limited module of ORYZA200 0 in Morogoro region of Tanzania. Materials and Methods Field Experiment A field study was conducted at Sokoine University of Agriculture (SUA), Morogoro, in 2012 2013 to evaluate the effects of several water management alternatives for the traditional continuously floode d rice system (CF) and system of rice intensification ( SRI ) on crop yield. The study was conducted for two seasons, the dry season (Oct ober 2012 Feb ruary 2013) and the wet season (Feb ruary 2013 Aug ust 2013). A randomized block design was used in thre e replicates of lowland rice with four water treatments: (i) continuous flooding (CF), (ii) intermittent wetting and drying (AWD) irrigation as applied to traditional SRI and (iii) two treatments with AWD and 50 and 80% irrigation volume of SRI Transplan ting included 1 plant per hill at 25 cm X 25 cm spacing, Crop, s oil water and weather parameters needed for the ORYZA200 model were measure d during the experiment (Chapter 2). Crop parameters included shoot biomass, leaf area index (LAI), leaf nitrogen ( N) and crop yield. Crop samples were taken during the growing season from eight hills to determine specific leaf area (SLA) and biomass of
58 green leaves, dead leaves, stems, and panicles. Soil water parameters measured were field capacity (%vol.), permanent wilting point (%vol.), saturation (%vol.) and saturat ed hydraulic conductivity (mm/hr.). Water inputs irrigation (m 3 ) and rainfall (m m ) were measured by a propeller type flow meter and a graduated rain gauge respectively R ainfall and other weather para meters were also collected at an automat ic weather station within the SUA campus and data were compared with the nearby Tanzania Meteorological Agency weather station ( also located within SUA campus ) W ater balance components (deep percolation, soil moistu re change ) were calculated from measured soil moisture at the depths of 10, 20, 30 60 80 and 90 cm from the soil surface Evapotranspiration (ET) was calculated as the residual term in the water balance equation. Further details can be found in Chapter 2 Model Setup, Description and Parameterization Model setup The ORYZA2000 (V 2.13, released in 2009) rice growth model was used in conjunction with the field experimental data (Chapter 2) were used to evaluate its performance. The emergence date was set a t the 280 th day of the year (DOY) with seedbed interval of 12 days and plant density of 16 hills m 2 with one plant per hill, similar to the field experiments (Chapter 2). Different input data, output results, controls and simulation mode were declared in different files in the model (Table 3 1). The model was run with water limited and non limiting nitrogen environment. The soil properties at the field experiment site were used to derive the soil parameters in the model. The weather data from the Tanzani a Meteorological Agency weather station for Morogoro w as used.
59 Model Description Crop Growth Replicated field data (Ch apter 2) were used to calibrate and validate the ORYZA2000 (V 2.13, released in 2009) model. The model considers four phe nological phases juvenile phase starting from emergence development stage to the vegetative stage (photoperiod sensitive stage) to panicle initiation stage, post panicle to 50% flowering stage and from flowering stage to the grain filling stage and finally to the maturit y stage. The corresponding development rates were set for different phenological phases (Table 3 2). Through the model interface input dat a files were defined with their specific data requirements C rop growth was simulated for each treatment with specifi c irrigation water regimes Details on the seedbed, soil hydraulic properties, applied N seedling per hill daily weather data and emergence dates were the primary input data for the ORYZA200 0 The model uses vertical distribution of leaf surface area to determine the amount of light profile within the canopy. Mean daily temperature is used to calculate the leaf area when the canopy is not closed. When the canopy closes, the increase in leaf area is determined from the increase in the leaf weight using th e specific leaf area (SLA) (Table 3 3) Drought affects leaf growth, leaf rolling, leaf senescence, photosynthesis, assimilate partitioning, root growth, and spikelet unfruitfulness. The transition from the exponential to linear growth phase is smoothened by taking weighted values of leaf area growth rates using the exponential and the linear equations. The dry matter
60 accumulation is calculated after subtracting the maintenance and respiration requirements. The biomass produced is allocated to the various p lant organs (i.e. leaves, stem, and panicles) as a function of phenological development which is calculated as a function of mean air temperature. Soil water dynamics Soil water dynamics in the ORYZA 2000 model were simulated for three soil types, poorly d rained lowland, regular upland, and well drained upland. Under ponded conditions, the model calculates ET using one of the three methods ( FAO Penman Montieth Priestley and Taylor, and Makkink), depending on the available meteorological data I n this study FAO Penman Montieth method was used because there was sufficient meteorological data The daily ET comprising of E vaporation (E) and T ranspiration (T) was obtained from the ponded layer The soil water dynamics w as simulated using the PADDY module within ORYZA2000 model (Wopereis et al., 1996; Bouman et al., 2001) which simulates the soil water dynamics for puddled lowland soils. Under non ponded conditions, the evaporation ( E ) was observed from the top soil layer while transpiration ( T ) was observed f rom different layers within the root zone. The deep percolati on losses from puddled layers wer e calculated using hydraulic properties of plo w sole and the non puddled soil beneath it. The maximum daily deep percolation loss is ponded water minus the ET los ses. When the water balance ( ( irrigation + rainfall (ET + deep percolation) ) depth exceeds the bund height, the excess water is assumed to be lost as runoff F or this study runoff did not occur during the dry season. When no percolation was computed (e. g. during the wetting and drying), inter layer fluxes
61 resulting from redistribution of water in the soil profile contributed to drainage For each layer when the total water input exce eds field capacity, the water was drained out at the rate that equals th e saturated hydraulic conductivity (Ks) (Table 3 2) If the drainage rate was less than water input, the water content could reach saturation and is allowed to develop a perched water table. If ground water was present in the soil horizon, the layer in the subsoil will drain until field capacity is attained. If there was no free drainage, it implies that there was restriction of water flow in one or more layer s and if the outflow was very low, excess water was redistributed upward to cause surface ponding (A rora, 2006). The rate of percolation from the puddled soil layer can either be fixed as a constant value or can be determined through an iterative process by using hydraulic properties of the plow sole and the non puddled subsoil. Under conditions of water limited production, the growth and development of the crop s are affected by drought. In ORYZA2000, the following effects of drought are taken into account: leaf rolling, spikelet sterility, reduced leaf expansion rate, changed assimilate partitioning, inc reased root depth, delayed vegetative development, increased leaf senescence, and decreased photosynthesis rate. For each of these processes, drought stress factors are calculated by water stress subroutine (WSTRESS) of ORYZA2000. The drought stress factor s are calculated from the soil water tension simulated with the soil water balance PADDY module within ORYZA2000 model ( IRRI.1993). Parameterization In ORYZA2000 model most of the crop parameters for rice are generic and may be used for all varieties (Bou man et al., 2001). Bouman et al. (2001) determined that the
62 following parameters from IR72 could be used as the standard parameters and then adjusted for the rice variety environment of interest (SARO 5 for this study). The parameters that are important in influencing simulated crop yield in ORYZA2000 are crop development rates, specific leaf area, leaf death rate, and fraction of stem reserves Therefore these parameters were adjusted during the calibration process to obtain the maximum yield from the model and compared to crop yield from the field experiments for different irrigation water regimes. ORYZA2000 was parameterized for the rice cultivar SARO 5 C rop phenology parameters ( leaf stem and panicle) for the dates of emergence, flowering, panicle initi ation, and physiological maturity for each treatment for the dry and wet seasons measured in the field study (Chapter 2) were used in modeling (Table 3 3) Model evaluation Model evaluation was achieved by comparing the measured and predicted values durin g the calibration (dry season) and the validation (wet season) T he ORYZA2000 model was calibrated and validated using experimental data (Chapter 2) D ata from the dry season (October 2012 to January 2013) were used for model calibration (Table 3 2 ) and (T able 3 3 ). The model was validated using separate data set f rom the wet season (March 2013 to June 2013) (Table 3 4 ). Model calibration and validation A combination of graphical and statistical techniques was used for evaluating model performance The mode l was calibrated using the dry season data (October 2012 to January 2013) while the model was validated for the four treatments for the wet season (March 2013 to June 2013) S catter plot s of simulated versus measured values were
63 developed for the soil wate r content, ponded depth, LAI, biomass, and yield. M easures used to evaluate model performance included coefficient of determination R 2 Nash Sutcliffe efficiency (NSE), percent bias (PBIAS), t he absolute (RMSEa) and normalized (RMSEn) root mean square erro rs. All statistical analyses were conducted using Statistical Analysis Software ( SAS ) (SAS, 1990). The data were analyzed using Tukey K ramer test. It compared the average effects of the four treatments. Tukey K ramer is a multiple comparison procedure for t esting pairwise differences of mean s (Somerville, 1993). Root Mean Square Error The absolute (RMSEa) and normalized (RMSEn) root mean square error values were calculated as : 0.5 ( 3 1) 0.5 ( 3 2) Where; n = n umber of observations RMSE a = a bsolute root mean square error (standard errors of measure d va lues ), RMSE n = n ormalized root mean square error (coefficient of variation of measured value s ),
64 Y i and X i = s imulated and measured values respectively, where is the mean of all measured values Nash Sutcliffe efficiency (NSE) The Nash Sutcliffe efficiency (NSE) i s a normalized statistical tool used to measured data variance (Nash and Sutcliffe, 1970). The NSE indicates how well the plot of observed versus simulated data fits the 1:1 line. The NSE can range from with 1 corresponding to a perfect match of model predictions to the observed data. A NSE of 0 indicates that the model predictions are as accurate as the mean of the observed data. A NSE of less than 0 indicates the ob served mean is a better predictor than the model. Moria si et al. (200 7 ) categorized the model performance as follows: , The NSE statist ic is calculated as : ] ( 3 3) Where; Y i obs = observ ed value for the constituent being evaluated Y i sim = simulated value for the cons tituent being evaluated Y mean = mean of observed data for the cons tituent being evaluated n = total number of observations.
65 Percent bias (PBIAS) The PBIAS meas ures the average trend of the simulated data determining if they are larger or smaller than their observed counterparts (Gupta et al., 1999). Positive values indicate model under estimation bias, and negative values indicate model over estimation bias (Gup ta et al., 1999). PBIAS is calculated using the following equation : ( 3 4) Where; Y i obs = observ ed value Y i sim = simulated value PBIAS = deviation of constituent being evaluated (%) Water Management Scenarios Several water management scenarios were analyzed using ORYZA2000. The water management scenarios included different fractions of irrigati on volume applied for the traditional SRI and a rainfed scenario with no irrigation. The scenarios included 90% SRI, 85% SRI, 80% SRI, 75% SRI, 70% SRI, 50% SRI, and rainfed rice. Simulations were run from the day of transplanting to physiological maturity for each season. Results from each scenario were used to identify the optimum irrigation management strategy that resulted in maximum rice yield.
66 Results and Discussion Model Performance Calibration For both seasons, the yields were more than actual Ta nzania rice yield (6 8 tons/ha) (URT, 2000) The model was calibrated using data for the dry season Parameters used in the calibration process were crop development rate, specific leaf area, leaf death rate and fraction of stem reserves The most influe ntial parameters were crop development rate and leaf death rate. T he model simulated well the crop yield indicators (biomass, LAI, number of tillers and plant height) for different irrigation treatments (Table 3 5) The RMSE for the yield varied from 30 kg /ha for the 50% SRI and 157 kg/ha for the CF treatments (Table 3 5) The NSE values for the yield for the four treatments ranged from 0.69 (CF) to 0.89 (50% SRI and SRI), an indication that the model performed well in simulating the most important determin ant of crop performance. The comparison between measured and simulated yield s for all four treatments for the dry season showed under prediction with all points below the 1:1 line (Fig ure 3 1). The RMSE for the total biomass ranged from 174 kg/ha for 80% SRI to 1259 kg/ha for CF treatment. Simulated and measured biomass of individual plant organs were also compared (Fig ure 3 3). The RMSE for the LAI ranged from 0. 51 to 0.91 for the SRI and 5 0% SRI system (Fig ure 3 4) For all treatments, t he NSE for total biomass ranged from 0.86 (80% SRI) t o 0.9 4 (CF) while it ranged from 0.64 (CF) to 0.85 (50% SRI) for the LAI (Table 3 5 ) H igh NSE values indicated that model performed well in simulating crop growth indicators during the calibration period Using the crit eria of Moriasi et al (2007), all NSE values were more than 0.6 indicat ing that the model
67 The model performance was satisfactory in predicti ng crop yield and its indicators for different irrigation trea tments. The model show ed increasingly better performance from f looded treatment to reduced irrigation treatment s This mean s model work s better under non flooded conditions Overall results for the calibration period (dry season) show that ORYZA200 0 is wel l suit ed for simulating the effects of water on crop growth performance measures as well as yield for a wide range of irrigation management for rice production in Tanzanian conditions. Validation The m odel predict ed the yield and all crop growth paramete rs well for the validation period (wet season) Comparison between simulate d and measure d above ground biomass for different treatments for the validation period are presented in Fig ure 3 2 The RMSE for the total biomass ranged from 1119 kg/ha for 80% SRI to 1420 kg/ha for 50% SRI treatments. The RMS E for the LAI ranged from 0.5 for SRI to 1.9 for the 50% SRI treatment The NSE for biomass predictions ranged from 0.6 0 (80% SRI) to 0.83 (SRI) while it ranged from 0.6 0 (SRI) to 0.72 (80% SRI) for the LAI ( Fi g ure 3 4 ) T his indicated that the model perform ance was good in simulating crop growth indicators during the validation period The RMSE for the yield varied from 129 kg/ha for the 50% SRI to 182 kg/ha for the SRI treatments. The NSE values for yield for the four treatments ranged from 0.6 (CF) to 0.83 (80% SRI) (Table 3 6 ). Using the criteria of Moriasi et al (200 7 ), all NSE values > 0.6 indicated that the model performance ranged very Overall, the calibration and the valid ation results show that model performed satisfactorily in simulating plant growth as well as yield
68 under a diversity of soil moisture (Fig ure 3 5) conditions and ponding depth (Fig ure 3 6) for the all four treatments for both seasons. Evaluation of Water Management Scenarios ORYZA2000 model was run for several water management scenarios starting from 50% SRI to 90% SRI. Results of these scenarios were used to find the range whe re the yield plateau occurred. This plateau occurred between 70 and 80% for bo th season s, thus the model was run with 1% increment s to find the specific regime which resulted in highest yield. Yield increased until 70% SRI for the wet season and 75% SRI for the dry season after which it started to drop This means that the optimum i rrigation strategies for the wet and dry seasons were 70% SRI and 75% SRI, respectively (Fig ure 3 7). Both ET and deep percolation water losses increased with an increase in irrigation ( Table 3 7 ) and (Table. 3 8 ). Similar observations have been made by Bo uman et al. ( 200 4) and Kukal and Aggarwal (2002). T he 90% SRI water regime was at saturation point below root zone (60 cm) from flowering stage to senescence indicat ing higher deep percolation losses. The soil moisture at 60 cm depth for the 75% SRI was 4 3% at transplanting to panicle initiation near field capacity (Table 2 2) but low er than saturation compared to 45 % for the SRI indicating lower deep percolation losses Simulated water balance components for the dry and wet se asons are presented in Tabl e 3 7 and Table 3 8 respectively. For dry season, t he drainage volume losses from CF and SRI were 1037% ( 5110 mm ) and 38% ( 185 mm ) more than that for 75% SRI respectively ; the ET losses for CF and SRI were 34% ( 229 mm ) and 3% ( 19 mm ) more than that for 75 % ET losses were higher for CF and SRI than that for 75% SRI because of excessive evaporation (Table
69 3 7 ) The drainage volume losses for CF and SRI were 18% (155 mm) and 16 % (134mm) higher than that for 70% SRI respectively. The ET from CF and SRI were 2 3% (132 mm) and 8% (44 mm) higher than ET from 70% SRI respectively. W ater savings (irrigation volume ) from 70% SRI was 334 mm and 195 mm compared to the CF and SRI respectively. The savings (irrigation volume ) was 16% ( 71 mm ) when compare the 70% SRI ov er the 80% SRI which was determined to be the best from experimental data analyses (Table 3 8) Results show that by reducing the irrigation volume for the traditional SRI by 25% (75% SRI) for dry se ason, additional yield of 2534 kg/ha can be obtained The water use efficiency (WUE) also improved from 429 kg/m 3 for 75%SRI to 255 kg/m 3 for SRI For the wet season, t he yield gain from the 70% SRI was 547 kg/ha higher c o mpared to the traditional SRI while the WUE increased from 231kg/m 3 for 70% SRI to 361 kg/ m 3 for the traditional SRI. The 70% SRI also showed a 30% reduction in irrigation volume Only 369 mm of supplementary irrigation was needed for 70% SRI b ecause of significant rainfall during the wet season (Fig ure 2 2 ) Conclusions ORYZA2000 was able to a ccurately simulate the rice production system under a variety of water availability conditions ranging from flooded to different levels of AWD irrigation management used with SRI. For the dry season, the model performance was good to very good in simulatin g plant growth ( biomass NSE = 0.86 to 0.9 4 ; LAI NSE = 0.64 to 0.85 ). The model performance in simulating yield was satisfactory to very good ( NSE = 0.60 to 0.80) Simul ation results showed that 75% SRI is the optimum irrigation strategy for the dry season while it was 70% SRI for the wet season On per unit area basis 75% SRI for the dry season and 70% SRI for the wet season res ulted in
70 additional yields of 2, 662 kg/ha and 747 kg/ha, respectively compared to the SRI and save d 35 % and 19% of water respect ively Due to the fact that rainfall, other climatic factors, soils and crop factors may vary from year to year long term simulations will be needed to determine the long term yield benefits. Results from this study show ed significant water savings from 70 and 75% SRI which can be used for other beneficial purposes such as environmental flows, human consumption, power generation increase in agricultural production area and industrial use.
71 Figure 3 1 S imulated against measured g rain yield for the four irrigation treatments for the calibration period ( Dry season, October 2012 to January 2013 ) 6000 7000 8000 9000 10000 11000 12000 6000 7000 8000 9000 10000 11000 12000 Simulated yield (kg/ha) Measure yield (kg/ha)
72 Figure 3 2 S imulated against measured total shoot biomass for the four irrigation treatments; (a) CF is continuously flooded rice, (b) SRI is System of Rice Intensification, (c) 80%SRI and (d) 50% SRI are treatments with 80% and 50% of irrigation volume compared to SRI between transplant a n d panicle initiation for the valid ation period ( wet season March 201 3 to June 2013) 0 4000 8000 12000 16000 0 4000 8000 12000 16000 Simulated biomass (kg/ha) Mesured biomass (kg/ha) 0 4000 8000 12000 16000 0 4000 8000 12000 16000 Simulated biomass (kg/ha) Measured biomass (kg/ha) 0 4000 8000 12000 16000 0 4000 8000 12000 16000 Simulated biomss (kg/ha) Mesured biomass (kg/ha) 0 4000 8000 12000 16000 0 4000 8000 12000 16000 Simulated biomass (kg/ha) Measured biomass (kg/ha)
73 F igure 3 3 Simulated and measured biomass for different plant organs for the four irrigation treatments against days after transplanting (DAT ); (a) CF is continuously flooded rice, (b) SRI is System of Rice Intensification, (c) 80%SRI and (d) 50% SRI are t reatments with 80% and 50% of irrigation volume compared to SRI between transplant and panicle initiation for the calibration period ( D ry season October 2012 to January 2013) 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 0 10 20 30 40 50 60 70 80 90 Biomass (Kg ha 1) DAT (a) panicle simulated stem simulated leaves simulated Panicle measured stem measured leaves measured 0 1000 2000 3000 4000 5000 6000 0 10 20 30 40 50 60 70 80 90 Biomass (Kg /ha) DAT (b) Stem simulated Leaves sim Panicle simulated Stem measured Leaves measured Panicle measured 0 1,000 2,000 3,000 4,000 5,000 6,000 0 10 20 30 40 50 60 70 80 90 Biomass (Kg ha 1) DAT (c) Panicle Simulated stem simulated leaves simulated Panicle Measured Stem measured leaves measured 0 1,000 2,000 3,000 4,000 5,000 0 10 20 30 40 50 60 70 80 90 Biomass (kg ha 1) DAT (d) Stem sm Leaves simulated Panicle simulated Stem measured Leaves measuted Panicle measured
74 Figure 3 4 Simulated and measured LAI of four different w ater treatments against days after transplanting (DAT); (a) CF is continuously flooded rice, (b) SRI is S ystem of Rice Intensification, (c) 8 0%SRI and (d) 5 0% SRI are treatments with 50% and 80% of irrigation volume compared to SRI between transplant a n d p anicle initiation for the calibration period ( D ry season October 2012 to January 2013) 0 2 4 6 8 10 0 20 40 60 80 100 LAI DAT a Simulated Mesured 0 2 4 6 8 0 20 40 60 80 100 LAI DAT b Simulated Mesured 0 2 4 6 8 0 20 40 60 80 100 LAI DAT c Simulated Measured 0 2 4 6 8 0 20 40 60 80 100 LAI DAT d Simulated Measured
75 F igure 3 5 Simulated against measured soil water content at 20cm depth from soil surface for different days after transplant (DAT) for the calibration period ; (a) CF is continuously flooded rice, (b) SRI is System of Rice Intensification, (c) 8 0%SRI and (d) 50% SRI are treatments with 8 0% and 5 0% of irrigation volume compared to SRI between transplant a n d panicle initiation (wet season March to August, 2013) 47.8 48.3 48.8 49.3 0 10 20 30 40 50 60 70 Soil moisture (%vol.) DAT a CF Measured CF Simulated 0 20 40 60 0 10 20 30 40 50 60 70 Soil moisture (%vol.) DAT b SRI Measured SRI Simulated 0 20 40 60 0 10 20 30 40 50 60 70 Soil moisture (% vol.) DAT c 80% SRI Measured 80% SRI Simulated 0 20 40 60 0 20 40 60 80 Soil moisture (% vol.) DAT d 50% SRI Measured 50% Simulated
76 Figure 3 6 Simulated and measured ponding depth for different days after transplant (DAT) for the validation period (wet season March 2013 to August 2013). 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 90 100 Ponding Depth (mm) DAT Simulated Measured
77 Figure 3 7 Simulated yield for different irrigation water regimes/ scenarios 0 2000 4000 6000 8000 10000 12000 14000 50 60 70 80 90 100 110 Simulated Rice Yield (kg/ha) Percentage of irrigation water applied for SRI Dry season simulated yield Wet season simulated yield
78 Table 3 1 Input data files and descriptions for the ORYZA2000 model File Descriptions Control data file The directories and pathways for the different input and output file Soil data file Specific soil information to be used in the soil and water sub models. Experimental data file Simulation options, modes of running, dates, cultivation method, condition of simulation (for this case water limited), transplant or direct seeding, irrigation parameters, Nitrogen parameters. Crop data file Physiological and phenological parameters correspondent to a specific rice variety. Weather data file FORTRAN simulation environment (FSE) format containing weather data for the experimental duration. Reruns data file Programming method to do multi ple simulation runs
79 Table 3 2 The c alibrated soil parameters for ORYZ2000 Soil parameters Depth from the soil surface (cm) Values Matric bulk density (g/m3) 0 25 1.45 25 50 1.36 50 75 1.35 75 100 1.33 100 125 1.31 125 150 1.33 Field capacity (%vol ) 0 25 40.1 25 50 40.7 50 75 40.8 75 100 40.9 100 125 41.1 125 150 40.9 Permanent wilting point (%vol.) 0 25 28.7 25 50 28.3 50 75 28.2 75 100 28.2 100 125 28.1 125 150 28.2
80 Table 3 2 Continued Soil parameters Depth from the soil surface (cm) Values Saturated hydraulic conductivity (mm/hr.) 0 25 0.35 25 50 1.18 50 75 1.34 75 100 1.68 100 125 2.09 125 150 1.68 Base saturation (% vol.) 0 25 45.3 25 50 48.5 50 75 48.9 75 100 49.8 100 125 50.6 125 150 27.9 Total acidity 0 25 0.1 25 50 0.2 50 75 0.2 75 100 0.2 100 125 0.3 125 150 0.9
81 Table 3 3 The c alibrated development rate constant (DRC) (C day 1 ) for the dry season G rowth Development Stage Value Juvenile phase 0.0006 Photoperiod sensitive phase 0.0007 Panicle development phase 0.0014 Reproductive phase 0.0010 Table 3 4 Validated leaf area index (LAI) for the w et season Developmen t stages (DVS) LAI v alues for four treatment CF SRI 80% SRI 50% SRI Transplanting 0.65 0.65 0.65 0.65 Vegetative 2.19 2.31 2.35 2.08 Panicle initiation 6.34 6.93 7.39 6.37 Flowering 7.77 8.99 9.03 7.29 Senescence 2.73 2.95 3.38 2.85 Maturity 2.02 2 .22 2.67 2.38 CF is continuously flooded rice, SRI is System of Rice Intensification, 50%SRI and 80% SRI are treatments with 50% and 80% of irrigation volume compared to SRI between transplant a n d panicle initiation stage
82 Table 3 5 Model calibration results for simulations of crop growth and yield va riables (D ry season Oct ober 2012 January 2013) Treatments Variables Simulated Measured R 2 NSE PBIAS RMSE 50% SRI Leaf Area Index 3.31 3.54 0.89 0.85 9.3 0.9 Total B iomass (kg/ha) 4388 4634 0.94 0.91 4.5 989.7 Yield (kg/ha) 7578 7482 0.94 0.89 0.3 30.3 80% SRI Leaf Area Index 3.57 3.55 0.90 0.80 6.0 0.6 Total Biomass (kg/ha) 6120 6198 0.97 0.86 2.6 934. 7 Yield (kg/ha) 11546 11452 0.9 8 0.84 0.3 36. 9 SRI Leaf Area Index 3.93 3.55 0.78 0.71 3.3 0.5 Total Biomass (kg/ha) 6198 6253 0.97 0.87 1.6 174. 3 Yield (kg/ha) 9551 9680 0.92 0.89 0.7 84. 5 CF Leaf Area Index 3.53 3.61 0.75 0.64 12.2 0.6 Total Biomass (kg/ha) 4037 4763 0.9 6 0.94 17.3 1259 .1 Yiel d (kg/ha) 8222 8689 0.83 0.69 1.6 157.1 CF is continuously flooded rice, SRI is System of Rice Intensification, 50%SRI and 80% SRI are treatments with 50% and 80% of irrigation volume compared to SRI between transplant ad panicle initiat ion
83 Table 3 6 Model validation results for simulations of crop growth and yield variables for the valida tion period ( W et season February 2013 June 2013) Treatment Variables Simulated Measured R 2 NSE PBIAS RMSE 50% SRI Leaf Area In dex 3.62 3.2 0 0.93 0.7 0 18. 3 1.9 Total Biomass (Kg/ha) 6581 5714 0.79 0.62 16.9 1440 .2 Yield (Kg/ha) 4285 4992 0.94 0.67 1.5 128.5 80% SRI Leaf Area Index 4.05 3.77 0.94 0.72 2.0 1.5 Total Biomass (Kg/ha) 7582 7377 0.93 0.6 0 4.9 1119 .1 Yield (Kg /ha) 5027 6009 0.91 0.83 2.9 129 .2 SRI Leaf Area Index 3.78 3.56 0.92 0.6 0 2. 9 0.5 Total Biomass (Kg/ha) 7193 7097 0.85 0.83 3. 3 1273 .1 Yield ( Kg/ha) 5001 5994 0.85 0.65 5.9 182 .3 CF Leaf Area Index 3.89 3.23 0.96 0.69 5. 4 0.8 Total Biomass (Kg/ ha) 7468 6255 0.76 0.73 9. 5 1254 .1 Yield (Kg/ha) 4937 5644 0.96 0.6 0 11.4 142.1 CF is c ontinuous ly f lood ed rice SRI is System of Rice Intensification, 50%SRI and 80% SRI are treatments with 50% and 80% of irrigation volume compared to SRI between t ra nsplant ad panicle initiation
84 Table 3 7 Simulated evapotranspiration (ET), drainage, and change in soil water storage along with irrigation and rainfall for the calibration period ( D ry season Oct ober 2012 to January 2013 ) T reatments Rainfall (mm) Irrigation (mm) ET (mm) Drainage (mm) Change in Storage (mm) CF 489 9300 908 5603 3277 SRI 489 1980 698 678 1093 90% SRI 489 1782 681 520 1070 80% SRI 489 1694 686 538 959 75% SRI 489 1485 679 493 802 70% SRI 489 1386 642 45 3 780 50% SRI 489 1040 639 380 510 CF is continuously flooded rice, SRI is System of Rice Intensification, 50%SRI and 80% SRI are treatments with 50% and 80% of irrigation volume compared to SRI between t ransplant ad panicle initiation Table 3 8 Simulated evapotranspiration (ET), drainage, and change in soil water storage along with irrigation and rainfall for the validation period ( W et season February 201 3 to June 2013 ) Treatments Irrigation (mm) Rainfall (mm) ET (mm) Drainage (mm) change in storage (mm) CF 787 1419 694 1019 493 SRI 648 1419 606 998 463 90% SRI 604 1419 601 962 460 80% SRI 524 1419 596 895 452 70% SRI 453 1419 562 864 446 50% SRI 322 1419 542 791 408 CF is continuously flooded rice, SRI is Sy stem of Rice Intensification, 50%SRI and 80% SRI are treatments with 50% and 80% of irrigation volume compared to SRI between transplant ad panicle initiation
85 C HAPTER 4 CONCLUSIONS AND RECOMMENDATIONS C onclusions Field Experiment Reducing irrigati on for rice production compared to the traditionally practiced continuous flooding (CF) facilitates soil aeration, increases microbial activities and promotes strong root growth leading to more number of tillers and yields as well as increased water produc tivity. These are the findings of a field experiment conducted to evaluate three different levels of irrigation input (SRI, 80% SRI, and 50% SRI) with the SRI production system against the traditionally practiced continuous flooding (CF) for the two (wet a nd dry) rice growing seasons in Tanzania. The conventional SRI and 80% SRI treatments outperformed the CF with regards to above ground biomass, number of tillers, and LAI for both seasons. The 80% SRI treatment produced 2.76 and 0.36 tons/ha more yield tha n CF with 456 and 263 mm less irrigation volume for the dry and wet season s respectively. The water use efficiency (WUE) for the 80% SRI was highest and was 60 and 100 % higher than CF for the wet and dry seasons respectively Under limited water availab ility scenario, applying 50% of the water needed for SRI is still a viable alternative, especially in the wet season during which the yield loss was only 14% compared to the high irrigation input treatment CF. Additionally, 50% SRI saved 54% of water compa red to CF This water saving could be allocated to other agricultural commodities Overall, 80% SRI was the optimum irrigation strategy with respect to crop yield as well as water productivity because it increases the yield by 24% and 6 % for the dry and
86 we t seasons but uses 65% and 52 % less water, respectively. On annual basis (two crops per year), 80% SRI and conventional SRI can save 719 mm and 475 mm of water compared to conventional CF. These water savings can be used for other purposes including irriga tion supplies for additional rice areas for achieving food security. Modeling Comparison of ORYZA2000 predictions with measurements from the experiments for the two seasons showed that the model simulated the effects of reductions in water input on rice y ield for Tanzanian conditions satisfactorily and can be used to determine sustainable irrigation management strategies. Model performance measures including the m ost widely used Nash Sutcliffe E fficiency (NSE), showed satisfactory to very good performance for the validation period, for growth indicators (Biomass: NSE = 0,60 to 0.94; LAI = 0.60 to 0.85) as well as yield (NSE = 0.60 to 0.89). When the effects on yield were evaluated for a variety of irrigation input scenarios for SRI practice (50% SRI to SRI) the best scenario s for the dry and wet seasons were 75% SRI and 70% SRI, respectively. For the dry season, implementation of 75%SRI resulted in annual water saving of 422,223 ha cm of water volume and achieved an additional production of 4.9 million ton s of rice in Tanzania. For the wet season, supplementary irrigation was found to be appropriate to fill the gap of dry days. However, these results may not represent long term yield advantage because of the variability in rainfall, soil, climate, and crop factors.
87 Recommendations In view of the great diversity in rice production systems due to varied local biophysical and socio economic conditions, 80% SRI may not always be applicable to all regions of Tanzania and elsewhere with similar environment. Regio n specific verification is likely to help achieve the full potential of results from this study. On farm participatory research will be required to introduce site specific adaptations and to expose farmers and extension agents to the practice of reduced ir rigation inputs (alternate wetting and drying) with SRI practice. For i rrigation schemes that depend on dams as a source of water, 75%SRI seem to be a better strategy for the dry season to make effective use of the limited storage. One of the requirements for the success of SRI practice demonstrated in this study is a sufficient degree of water control. For the farmers to be able to apply the required amount of water regularly and reliably, it need development of easy to understand rule based (e.g.3 days o f wetting and 5 days of drying) irrigation management from transplanting to panicle initiation stage This can be achieved by organizing local training programs for farmers to realize the full potential of SRI practice. To implement the SRI irrigation stra tegies, further improvements in irrigation infrastructure and management capacities will be needed This will give farmers and water managers the ability to utilize smaller but reliable amounts of water as needed for the sustenance of crops and beneficial soil organisms. There is also a need to rethink and strengthen capacities for research and extension programs and appropriate policy formulation for large scale adoption by rice growers of Tanzania. Given the satisfactory to good performance of ORYZ200 in this study for the two seasons, this model seem to be
88 adequate for evaluating not only different irrigation management alternatives but also for evaluating rice production under current and changed climatic conditions using long term climatic data and simu lations.
89 AP P END I X STATISTICAL PARAMETERS Table A 1 Statistical results for comparing LAI values for the dry season Measures CF SRI 80% SRI 50% SRI Mean 3.888 3.408 3.367 3.296 S tandard D eviation 3.214 2.681 2.619 2.514 Coefficient of Variation 0.82 7 0.787 0.778 0.763 P values 0.028 0.032 0.029 0.031 Table A 2 Statistical results for comparing LAI values for the wet season Table A 3. Statistical test results for total biomass for the wet season Measures CF SRI 80% SRI 50% SRI Mean (kg/ha) 5105 5969 6189 4530 Standard Deviation (kg/ha) 4652 5106 5239 4265 Coefficient of Variation 0.911 0.855 0.846 0.941 P v alues 0.032 0.024 0.421 0.045 Measures CF SRI 80% SRI 50% SRI Mean 3.2 3 3.5 6 3.77 3.21 S tandard Deviations 2.57 2.96 3.0 2 2.4 5 C oefficient of V aria ti on 0.79 0.83 0.80 0.76 P values 0.0 3 0.02 0.03 0.03
90 Table A 4 Statistical test results for total biomass for the dry season. Measures 50% SRI 80%SRI SRI CF Mean (kg/ha) 9811 11087 10092 9346 Standard Deviation (kg/ha) 29. 68 138.85 77.35 31.34 Coefficient of Variations 0.003 0.013 0.008 0.003 P values 0.002 0.021 0.005 0.002 Table A 5 Tukey Kramer grouping for treatments l east squares means (Alpha=0.05) with regards to crop yield for the dry season Treatments 80% SR I SRI CF 50% SRI Mean (kg/ha) 11452 9680 8689 7481 Standard Deviation (kg/ha) 326.1 309.6 322.1 250.4 Coefficient of Variation 0.028 0.032 0.037 0.033 P values 0.039 0.018 0.019 0.028
91 Table A 6 Tukey Kramer grouping for treatments l east squares m ean s (Alpha=0.05) with regards to crop y ield for the wet season Treatments 80% SRI SRI CF 50% SRI Mean (kg/ha) 6009 5994 5644 4992 Standard Deviation (kg/ha) 33.9 151.6 120.6 33.6 Coefficient of Variation 0.006 0.025 0.021 0.007 P values <0.01 0.06 0.06 <0.01 Table A 7. Tukey Kramer test results for crop yield differences for the dry season Treatments 50% SRI against 80% SRI 50% SRI against CF 50% SRI against SRI 80% SRI against CF 8 0 % SRI against SRI CF against SRI Estimate ( kg/ha) 3970 1207 2198 2762 1771 990 Degree freedom 6 6 6 6 6 6 t Value 16.01 4.87 8.87 11.14 7.15 4 P values <0.01 0.011 0.01 0.01 0.01 0.03
92 LIST OF RE FERENCE Aggarwal, P.K., Kropff, M.J., Cassman, K.G., Ten Berge, H.F.M., 1997. Simulating ge notypic strategies for increasing rice yield potential in irrigated tropical environments. Field crops research. 51, 5 17. Allen, R.G., Pereira, L.S., Raes, D., Smith, M. Crop evapotranspiration: guidelines for computing crop water requirements. Rome: FAO 1998. 301p. Irrigation and Drainage Paper, 56 Alvaro C., Rehdanz, K., Tol, R.S.J. 2010. The economic impact of more sustainable water use in agriculture: A computable general equilibrium analysis. Journal of hydrology. 384, (3 4) 292 305. Anbumozhi, V., Yamaji E ., Yabuchi T., 1998. Rice crop growth and yield as influenced by changes in ponding water depth, water regime and fertigation level. Agriculture water management. 37, 241 253. Arora, V.K., 2006. Application of a rice growth and water balance m odel in an irrigated semi arid subtropical environment. Agriculture water management, 8(3), 51 57. Balisacan, A.M ., Sebastian, L.S., 2006. Securing rice reducing poverty: Challenges and Policy Directions. SEARCA, Phil Rice and DA BAR. 352. Behrouz, M .F., Fatemeh, J., Sayed Farhad M ., Mohammad Reza Y., 2010. Effects of irrigation water management on yield and water use efficiency of rice in cracked paddy soils. AJCS. 4(3), 136 141. Bhuiyan i S.I., 1992. Water management in relation to crop production : Case study on rice. Outlook agriculture. 21(4), 293 299. Boggess, W.G Ritchie, J. T. 1988 Economic and risk analysis of irrigation decisions in humid regions. Production Agriculture 1, 116 122. Borell, A., Garside, A ., Shu, F.K., 1997. Improving e fficiency of water for irrigated rice in a semi arid tropical environment. Field crops research. 52, 231 248. Bouman, B.A.M., Kropff, M.J., Tuong, T.P., Wopereis, M.C.S., ten Berge, H.F.M., Van Laar, H.H., 2001. ORYZA2000: Modelling Lowland Rice. Internat ional rice research Wageningen, Netherlands. Bouman, B.A.M. Martin M., 2004. More Rice, Less Water Integrated Approaches for Increasing Water Productivity in Irrigated Ric e Based Systems in Asia. Proceedings of the 4th International crop science congress.
93 A.R., Visperas, R. M., 2005. Yield and water use for irrigated tropical aerobic rice systems. Agriculture water management. 74, 87 105. Bouman, B.A.M., Van Keulen, H., Van Laar, H.H., Rabbinge, R. 1996 The School of de wit crop growth simulation models. Agricultural systems. 5(2), 171 198. Bouman, B.A.M., van Laar, H.H., 2006. Description and evaluation of the rice growth model ORYZA20 00 under nitrogen limited conditions. Agricultural system s 87, 249 273. Calzadillaa, A., Rehdanza K, Richard, S.J.T., 2010. Economic impact of more sustainable water use in agriculture: A computable general equilibrium analysis. 384 (3 4), 292 305. Cha pagain, A. K., Hoekstra, A. Y ., Savenije, H. H. G., 2006. Water saving through international trade of agricultural products. Hydrology & Earth System Sciences. 10(3),455 468. Champagain, A.K. Hoekstra, A.Y., 2011. The blue, green and grey water footprint of rice from production and consumption perspectives. Ecological economy. 70, 749 758. Chaudhary, D.K., 2003. Effect of water regimes and NPK levels on mid duration rice. (Oryza sativa L). M.Sc. (Ag.). Thesis, RAU, Pusa, Bihar. Dakishinamurthi, C., Mich ael, A.M., Sri Mohan ., Pande. S.L., 1973. Report on the visit to Thailands and Philipines (on Rice Cultivation). Water Technology Centre, IARI, New Delhi, 1 53. Dawe, D., 2002. Changing structure of the world rice market. 27(4), 355 370. Dobermann, A., 2004. A critical assessment of the system of rice intensification (SRI). Agricultural Systems 79, 261 281. Drenth, H. Ten Berge, H.F.M. Riethoven, J.J.M., 1994. ORYZA simulation modules for potential and nitrogen limited rice production. SARP Research Proceedings, IRRI/AB DLO, Wageningen, Netherland. 223. Experimental station for irrigation team, 2009. Penelitian Irigasi Hemat Air pada Budidaya Padi dengan metode SRI di Petak Tersier, Tasikmalaya Buku 22, Seri Penelitian Irigasi Hemat Air pada Budidaya Padi dengan metode SRI, Balai Irigasi. FAO. 1998. Proceedings of 19th Session of the International Rice Commission, Cairo, Egypt, Rome, 56, 110 117.
94 Faustin, P.M., John A.B ., Patrick M., 2001. Domestic water supply, competition for water resources and I WRM in Tanzania: a review and discussion paper. Physics and Chemistry of the Earth, Parts A/B/C. 27(11 22), 919 926. Frank, R. Rijsberman., 2006. Water scarcity: Fact or fiction? Agricultural water management. 80, 5 22. Ga i ni, A., Rahman, A., Rustam D. Hengsdijk, H., 2002. Water management experiments in Indonesia. Paper presented in international symposium on water wise rice production, IARI, New Delhi 2(3), 29 37. Guerra, L.C., Bhuiyan, S.I., Tuong, T.P., Barker, R., 1998. Producing more rice with les s water from irrigated systems. SWIM Paper 5. IWMI/IRRI, Colombo, Sri Lanka. 24. Gupta, R. Acharya C.L., 1993. Effect of mulch induced hydrothermal regime on root growth, water use efficiency, yield and quality of strawberry. J. Indian Soc. Soil Science. 41 (1), 17 25. Gupta, H.V., Sorooshian, S., P. O. Yapo., 1999. Status of automatic calibration for hydrologic models: Comparison with multilevel expert calibration. J. Hydrologic Engineering. 4(2), 135 143. Hafeez, M. M Bouman B.A.M Giesen V. N ., Vlek P., 2007. Scale effects on water use and water productivity in a rice based irrigation system (UPRIIS) in the Philippines. Agriculture water manages. 92, 81 89. Hijmans, R.J., Guiking Lens, I.M., Van Diepen, C.A., 1994. WOFOST, user guide for the WOF OST 6.0 crop growth simulation model. Technical document 12, DLO Winand Staring Centre, Wageningen, The Netherlands, 145. term plan for 1994 1998. IRRI, Los Banos, Philippines. Janssen, M ., Le nnartz B., 2007. Horizontal and vertical water and solute fluxes in paddy rice fields. Soil & Tillage research. 94, 133 141. Japhet J.K Rajabu K. Masolwa P., 2009. Freshwater management and climate change adaptation: Experiences from the Great Ruaha Ri ver catchment in Tanzania. 1(3), 220 228. Katambala Z., Kahimba F. C., Mahoo H. F., Mbungu W B., Mhenga F., Reuben P., Maugo M. Nyarubamba A., 2013. Adopting the system of rice intensification (SRI) in Tanzania. Agricultural sciences. 4(8), 369 375. Keu len, V.H ., Breman, H., 1990. Agricultural development in the West African Sahelian region: a cure against land hunger ? Agriculture, Ecosystems & Environment. 32(3 4), 177 197.
95 Kitova, H. H., 2001. The development and management of hydropower resources Tan zania. H&E course at ICH Trondheim, Norway, 6. Kombe, E.E., 2012. The system of rice intensification (SRI) as a strategy for adapting to the effects of climate change and variability: MS thesis. Morogoro: Sokoine University of Agriculture, Department of A gricultural Engineering and Land Planning. Krishna, A., Biradarpatil, N.K ., Channappagoudar, B.B., 2008. Influence of system of rice intensification (SRI) cultivation on seed yield and quality. Agriculture sciences. 21(3), 369 372 Kropff, M.J, Van Laar, H.H. Mathews R.B., 1994a. ORYZA1, an ecophysiological model for irrigated rice production. SARP Research Proceedings AB DLO, Wageningen, the Netherlands, 110. Kropff, M.J., Matthews, R.B., Van Laar, H.H. Ten Berge, H.F.M., 1995. The rice model ORYZA1 a nd its testing. In: Matthews, R. B., Kropff, M. J., Bachelet, D., Van Laar, H.H. (Eds.), Modelling the Impact of Climate Change on Rice Production in Asia. CAB International, Wallingford, UK, 27 50. Kukal, S.S ., Aggarwal G.C., 2002. Percolation losses of water in relation to puddling intensity and depth in a sandy loam rice (Oryza sativa) field. Agriculture water manages. 57, 49 59. Li, Y.H., 2001. Research and practice of water saving irrigation for rice in China. In: Barker R, Li Y, Tuong TP, eds. Water saving irrigation for rice. Proceedings of an International Workshop, Wuhan, China. Colombo (Sri Lanka). International Water Management Institute, 135 144. Li, Y. Barker, R., 2004. Increasing water productivity for paddy irrigation in China. Paddy Water Environ. 2, 187 193. Maganga, F.P., 2003. Incorporating customary laws in implementation of IWRM: some insights from Rufiji River Basin, Tanzania. Physics and Chemistry of the Earth, Parts A/B/C. 28(20 27) 995 1000. Mahoo, H.F. Mkoga, Z.J., Kasele, S.S ., Igbadun, H.E., Hatibu, N., Rao, K.P.C. Practices. Discussion Paper on Comprehensive Assessment of Water Management in Agriculture IWMI, Colombo, Sri Lanka. Mao et al ., 1993. Environmental impact of water saving irrigation for rice production. Proceedings of environmentally sound water resource utilization. AIT Bangkok, Thailand, 143 148.
96 Mati, B. M., Wanjogu, R., Odongo, B., Home, P. G., 2011. Introduction of the Syst em of Rice Intensification in Kenya: experiences from Mwea Irrigation Scheme. Paddy and Water Environment. 9(1), 145 154. McDonald, A.J Hobbsb, P.R. Rihaa, S.J., 2006. Does the system of rice intensification McMennamy, J.A., O_Toole, J.C., 1983. RICE MOD: a physiologically based rice growth Moriasi, D. N., Arnold, J. G., Van Liew, M. W., Bingner, R. L., Harmel, R. D., Veith, T. L. 2007. Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Trans. ASABE Mostafazadeh Fard, B ., Malano H.M., 1996. Infiltration in cracking clay soils as affected by initial soil moisture content. Iran agriculture research. 15, 133 154. Mostafazadeh Fard. B Fatemeh J, S.F Mousavi and Yazdani M.R, 2010. Effects of irrigation water management on yield and water use efficiency of rice in cracked paddy soils. AJCS 4(3), 136 141 Mousavi, S.F, Yousefi Moghadam S., Mostafazadeh Fard, B., Hemmat A. Yazdani M.R., 2009. Effect of puddling intensity on physical propert ies of a silty clay soil under laboratory and field conditions. Paddy water environment. 7 (1), 45 54. Mwakalila, S., 2005. Water resource use in the Great Ruaha Basin of Tanzania. Physics and Chemistry of the Earth, Parts A/B/C. 30(11 16), 903 912. Nas h, J.E. Sutcliffe J.V., 1970. River flow forecasting through conceptual models: Part 1. Discussion of principles. Journal of Hydrology. 10(3), 282 290. Nelson, G.C. (2009). Agriculture and Climate Change. An Agenda for Negotiation in Copenhagen. 16(6). Noel, B.S., 2011. The power of imagination in transnational mobilities. Global studies in culture and power. 18(6), 576 598. Panda, S.C, Rath, B.S, Tripathy, R.K. Dash, B., 1997. Effect of water management practices on yield and nutrient uptake in the dr y season rice. Oryza. 34, 51 53. Parihar, S. S., 2004. Effect of crop establishment method, tillage, irrigation and nitrogen on production potential of rice (Oryza sativa) wheat (Triticum aestivum) cropping system. Indian. Journal of agronomy. 49(1), 1 5 Parsi N. M., Yazdani, M ., Razavipour, T., 2003. The study of water application efficiency of paddy fields (Case study, irrigation system network of Sepid Roud, Guilan). 11th Seminar of Iranian National Committee on Irrigation and Drainage, Tehran, 111 1 20.
97 Paul, N.S., 1993. On the conservatism of the Tukey Kramer multiple comparison procedure. University of central Florida, Orlando, Florida USA. Statistical and probability letter 16, 343 345. Penning de Vries, F.W.T., Jansen, D.M., Ten Berge, H.F.M., B akema, A. (Eds.), 1989. Simulation of ecophysiological processes of growth in several annual crops. Simulation Monographs. Pudoc, Wageningen, The Netherlands, 271. Peter, H.G., Smith, J.C. Heather, C., 2011. Water use efficiency and productivity: rethink ing the basin approach. Water international. 36(7), 784 798. Rejesus, R.M., Palis, F.G., Rodriguez, D.G.P., Lampayan, R.M., Bouman, B.A.M. 2011. Impact of the alternate wetting and drying (AWD) water saving irrigation technique: Evidence from rice produce rs in the Philippines. Food policy. 36(2), 280 288. Sakurai, T., 2006. Intensification of rain fed lowland rice production in West Africa: Present status and potential green revolution. Developing economies. 44, 232 251. Sander, J.Z ., Wim, G.M.B., 2004. Review of measured crop water productivity values for irrigated wheat, rice, cotton and maize. Agricultural water management. 69(2), 115 133. Shadrack, M., 2005. Water re source uses in the great ruaha basin of Tanzania. Physics and Chemistry of the Earth, Parts A/B/C. 30(11 16) 903 912. Shan, D. K. Land Sattelmacher B., 2002. The ground cover rice production system (GCRPS) a successful new approach to save water and incr ease nitrogen fertilizer efficiency? In: Bouman, Hengsdijk, Hardy, Bindraban, Tuong, Lafitte and Ladha (Eds.). Water Wise Rice Production. Los Banos (Philippines): IRRI, (in press). Singh, U., 1994. Nitrogen management strategies for lowland rice cropping systems. In: Proceedings of the international conference on fertilizer usage in the tropics (FERTROP). Malaysian society of soil science, Kuala Lumpur, Malaysia, 110 130. Somerville, P.N., 1993. On the conservatism of the Tukey Kramer multiple comparison procedure. Statistics and probability letters. 16(5), 343 345. Stoop, W Uphoff N. T ., Kassam, A., 2002. A review of agricultural research issues raised by the System of Rice Intensification (SRI) from Madagascar: opportunities for improving farming sys tems for resource poor farmers. Agriculture system. 71, 249 274. Sudhir Yadava, Tao Li b, E. Humphreysb, Gurjeet Gill, S.S. Kukalc., 2011. Evaluat ion and application of ORYZA2000 for irrigation scheduling of puddled transplanted rice in North West India. Field crops research 122, 104 117.
98 Tuong, T.P. & Bouman, B.A.M., 2003. Rice Production in Water scarce environments. International rice research i nstitute, Manila, Philippines. Tabbal, D.F, Bouman B.A.M, Bhuiyan S.I.., Sibayan, E.B Sattar, M.A., 2002. On farm strategies for reducing water input in irrigated rice; case studies in the Philippines. Agriculture water manages. 56(2), 93 112. Talpur, M.A., Ji Changying, S.A., Junejo, A.A. Tagar ., Ram, B.K., 2013. Effect of different water depths on growth and yield of rice crop. African Journal of Agricultural Research. 8(37), 4654 4659. Tom, F., Bruce, L. Makarius M., 2004. Managing water amongst co mpeting uses. The Usangu wetland in Tanzania. 53(3), 277 286. Toshiyuki, T., Shoji, M., Takahiro N., Akihiro O. Tatsuhiko S., 2006. Rice yield potential is closely related to crop growth rate during late reproductive period. Field crops research. 96, 328 335. URT. 1995a. Rapid water resources assessment. Main report, Ministry of water, energy and minerals, 1 URT. 2000. United Republic of Tanzania, Soil fertility initiative concept paper. Report No.00/081 CP URT. Van Ittersum, M.K., Leffelaar, P.A., Van Keulen, H.K., Bastiaans, M.J. Oudrian, L., 2003. On approaches and applications of the Wageningen crop models. European Journal of Agronomy. 18, 201 234. Vijayakumar, M., Ramesh, S., Chandrasekaran, B. Thiyagarajan, M., 2001. Effects of system of rice intensification (SRI) practices on yield attribute yield and water productivity of rice (Oryza sativa l.). Res J Agric Biological Sci. 2(6), 236 242. Willem, A.S, Norman, U. Amir, K., 2002. A review of agricultural research issues raised by the system of rice intensification (SRI) from Madagascar: opportunities for improving farming systems for resource poor farmers. Agricultural systems management. 71, 249 274. Willmott, C.J., 1984. Evaluation of model performance in physical geography. Spatial statisti cs and models. 2, 443 460. Wopereis, M.C.S., Bouman B.A.M., Tuong T.P., Ten Berge H.F.M., Kropff M.J., 1996. ORYZA_W: rice growth model for irrigated and rain fed environments, SARP Research Proceedings, IRRI/ABDLO, Wageningen, Netherlands, 159.
99 Wopere is, M.C.S., 1993. Quantifying the impact of soil and climate variability on rain fed rice production. Ph.D. Thesis, Wageningen agricultural university, Wageningen, Netherlands, 188. Xinyou, Y., Schapendonk A.D.H.C.M., Martin J.K., Marcel V.O. Prems. B., 2000. Generic equation for Nitrogen limited Leaf Area Index and its Application in Crop Growth Models for Predicting Leaf Senescence, Annals of Botany 85, 579 585. Zwart, S.J and Bastiaanssen, W.G.M. 2004. Review of measured crop water productivity value s for irrigated wheat, rice, cotton and maize. Agricultural w ater m anagement. 69(2), 115 133.
100 BIOGRAPHICAL SKETCH Stanslaus Terengia Materu was born in Moshi Kilimanjaro (Kiboriloni) in the Northern Tanzania. He attended Sima njiro Secondary School where he graduated from Secondary School in 2001 and then to Majengo Secondary where he completed his high school in May 2004. In August of 2005, he moved to Morogoro and successfully passed the entrance exam at Department of Agricul tural Engineering and Land Planning of the Sokoine University of Agriculture (SUA) where he spent four years and worked with Ministry of Agriculture of United Republic of Tanzania as an Agro engineer II then move SUA as Tutorial Assistant. In July 2011, Stanslaus obtained a scholarship from USAID/Feed the Future (iAGRI) project to come to University of Florida (USA) and biological engineering with a concentration in land and water engineering in May 2014. His skills include agricultural systems, water project management, crop modeling, hydrological modeling, computer programing. Stanslaus aims to become an expert in t he field of biological and agricultural systems and work in challenging environments to serve his country from his expertise.
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E6HJR3V4D_ML24D0 INGEST_TIME 2014-10-03T22:16:23Z PACKAGE UFE0046853_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC