Overview Objectives

The project is improving the management of water and N fertiliser to increase farm incomes, improve environmental quality and reduce N2O emissions from agriculture. The systems to be studied are irrigated maize and wheat cropping systems and intensive vegetable farms in the western Yellow River basin of northern China, and intensive irrigated pasture and maize, and rain-fed wheat systems in Australia.

Project Background and Objectives

In China irrigated cropping demands careful management of water resources and other inputs. Wheat and maize are the two main irrigated crops grown in western provinces. Water use efficiency in these areas is often low despite water being a critical resource. The intensive nature of the cropping that is practised demands the use of fertilisers. Nitrogen fertiliser is the main type used but, as with water, its use is often inefficient and wasteful. The combination of water used inefficiently with more fertiliser than is needed creates environmental problems, beginning with nitrogen-rich runoff. Volatilisation of ammonia contributes to nitrogen loss, in turn requiring more fertiliser use. This also results in greenhouse gas emissions.

A Water and Nitrogen Management Model (WNMM) developed in past ACIAR research should help determine changes needed to improve current management practices. By working with the model and its associated decision support system in two AusAID projects in Inner Mongolia and Hebei provinces, better practices will be identified and then disseminated to farmers through the existing project channels.

Progress Reports (Year 1, 2, 3 etc)

Most of the planned activities and milestones outlined in the project proposal for the first year (of a four year project) have been achieved, and summarized as follows:

In China:
1. A comprehensive experimental and data collection protocol for the project was developed and distributed to all the project members.
2. Four experimental sites were selected, (a) irrigated maize system in the western part of the Inner Mongolia Autonomous Region (IMAR) in collaboration with AusAID project of Alxa League Environmental Rehabilitation and Management, and (b) three sites in Shanxi province; the northern site at Yuci in the Taiyuan Basin, the southern site at Yongji near the edge of the North China Plain and the Hongtong site which lies in between. Both Yuci and Yongji sites were extensively instrumented to measure all key water and N fluxes.
3. Farmer socio-economic survey, field experiments for quantifying efficiencies of irrigation and N fertiliser using 15N labelled urea, and collection of soil and plant samples for N dynamics have been completed.
4. Sophisticated instruments have been installed at Yuci and Yongji, including micrometeorological Eddy correlation facilities, automatic soil moisture instruments and automated chambers with gas chromatography (GC) for N2O (Yuci site), and microplots receiving fertiliser enriched with stable 15N isotope to trace the fate of fertiliser N.
5. First year comprehensive water and N balance measurements have been completed at Yuci, most of the collected soil and plant samples have been analysed.
6. Targeted farmers socio-economic survey was completed in Fengqiu county to study the reasons for low adoption rates of water saving techniques recommended in a recently completed ACIAR project. Based on the survey data and the data from this previous ACIAR project, a conceptual module of the economic component has been established for water and nutrient management model (WNMM).
7. Chinese Academy of Sciences has approved 3 million RMB (A$500,000) co-funding to support this project (Prof Zheng Xunhua of Atmospheric Institute of Physics) to establish a state of art auto-chamber system for N2O and NOx measurements at the Yongji site.
8. The Shanxi Agricultural Comprehensive Development Office (SACDO) has agreed to fund 750,000 RMB (A$125,000) to implement the research outcomes of the project by establishing demonstration farms in all three counties in Shanxi.

In Australia:

The Australian component of project is complementary to projects funded through the CRC-Greenhouse Accounting, Australian Greenhouse Office (AGO) and GRDC, and mainly focus on model development.
9. The GIS interface of the WNMM has been converted to ArcGIS 9.x from ArcView GIS 3.x. The new WNMM-GIS interface for ArcGIS 9.x will make the spatially-referenced WNMM independent of the limitations set up by ESRI. However, the old version is still usable if the users have ArcView GIS 3.x.
10. The WNMM has been significantly modified and improved for Australian conditions:
A new component for simulating the localised soil pH change due to urea hydrolysis near urea granules and animal urine patches was added.
New crop growth models have been added to WNMM, including; SUCROS, ManageWheat model (John Angus, CSIRO), IRRI Oryza 2000 (to simulate rice growth), the CERES crop growth model (to simulate wheat and maize growth) and GrassGro (to simulate pasture growth and grazing). The new WNMM is a versatile agro-ecosystem model.
Three additional N2O emission modules have been added to WNMM for predicting the N2O emissions from soils; DNDC, DAYCENT and FASSET.
The web-based version of WNMM is nearly completed. This version of WNMM will allow users to access WNMM through internet connection to setup inputs files, to conduct simulations on our WNMM server, and to get the simulation results in text report and chart format produced by R software (http://www.R-Project.org) on our server. It is expected that the single location version will be accessible, rather than the spatially distributed version.

11. Applications WNMM in Australia and overseas.
The WNMM was adopted by CRC-GA for simulating water and N dynamics, and N2O emissions, for rain-fed wheat in Victoria and Western Australia and irrigated pastures in Victoria; N dynamics and NH3 volatilisation for south Korea, and water and N dynamics for an irrigated maize and wheat system in the Yaqui valley, Mexico by scientists from Stanford University, and legume systems in China by scientist from the Institute of Soil Science, Chinese Academy of Sciences. Significant outcomes are:
At Rutherglen, Victoria, the WNMM was applied to simulate N2O emissions from a rain-fed wheat cropping system with different stubble management practices, conventional cultivation of burnt stubble and direct drill with stubble retained. Both measured and simulated results showed that the differences in N2O emissions between the stubble management were small and the main driver for N2O emissions is soil moisture rather than the availability of soil mineral N.
The calibrated WNMM was used to simulate N2O emissions using historic climate data from 1968 to 2004 and four climate change scenarios (A1FI, A2, B1 and B2) of IPCC from 1990 to 2100 for seven scenarios of N fertilizer applications at the Rutherglen site. It was found that there were huge variations of annual N2O emissions, ranging from 0.1 to 0.6 kg N ha-1year-1 during the past 37 years. The annual N2O emissions were significantly correlated to climatic variables: annual average of daily maximum air temperature, annual rainfall and the N fertilizer application rate. The emission factors are around 0.1% of applied N which is well below the IPCC defaults value of 1.25%.
The water and N fluxes were successfully simulated by WNMM for the irrigated grazing pasture system at Kyabram. Four gas modules, DNDC, DAYCENT, WNMM and FASSET, under WNMM framework were compared for N2O simulations. The simulations of N2O emissions using the simplest WNMM module were the most accurate.

12. Soil N dynamics in irrigated maize systems impacted on by N and stubble management at Griffith, NSW were studied. The field measurements of denitrification, mineral N content and recovery of 15N labelled urea from micro-plots with and without ammonium thiosulfate (urease inhibitor), were complemented with laboratory studies of denitrification and N2O flux. Only 64-68% of applied labelled fertiliser was recovered from the 15N microplots. Significantly more fertiliser N was recovered in the grain from the stubble incorporated treatment than the stubble burned treatment and there was greater recovery of fertiliser N in the soil at the end of the experiment in the stubble burned treatment. The N2O emissions were strongly correlated with N application rates between 0 and 300 kg N/ha, with significantly more N2O emissions in the stubble burnt treatment than that the stubble incorporated treatment. The emission factors are between 1.5 to 2.5% of applied N which is higher than the IPCC’s default value in contrast to the rain-fed wheat system in Australia.

13. A prototype of a handheld-based agricultural decision support system has been developed for advising farmers about N fertiliser use for wheat cropping in Australia. The system is operated on a hand held computer with built in digital camera and wireless functions (wireless network or GPRS). The soil N availability has been found to be correlated to the shoot density at early stages or the ground vegetation fraction which can be calculated from the in-situ photos taken by the built-in digital camera. The wireless networks or GPRS is used to download the real time weather data to calculate the wheat growing stages using the simplified WNMM model. The concept and the developing protocol may be adoptable in the Northern China region to extend the agricultural research results by this ACIAR project to the farmers who are very keen to hear and to adopt BMPs for social, economic and environmental benefits.

Most of the milestones outlined in the project proposal for the second year (of this four year project) have been achieved, and are summarized as follows:

In China:
1. Field experiments on irrigated maize in the Inner Mongolia Autonomous Region (IMAR), in collaboration with the AusAid project “Alxa League Environmental Rehabilitation and Management” have been completed. Using a combination of field measurement, modelling and 15N tracer techniques we found that 25 to 40% of irrigation water, and 186 to 255 kg N ha-1 of nitrate, was leached below the root zone. It was estimated that 50-90% of applied N fertiliser was lost.
2. The water and N management model, WNMM, has been adapted for simulating water and N dynamics under maize cropping in IMAR. Best management practices for these systems have been identified on the basis of WNMM simulations.
3. Two years of field experiments on water and N dynamics, plant growth and yield have been completed on irrigated wheat and maize at Yongji and Hongtong (Shanxi Province). The data is being used for testing WNMM in these environments. The Yongji site is extensively instrumented, including an Eddy Covariance system for measuring evapotranspiration and CO2 fluxes and a wireless soil moisture monitoring system. The preliminary results indicate that in Shanxi 30 to 160 kg/ha N fertiliser can be saved without reducing maize yields, and the corresponding saving for wheat is 40 to 100 kg/ha. This translates to 120 to 640 RMB/ha cost saving (1A$=6RMB). Similarly, significant amounts of irrigation water can be saved without lowering yield.
4. Open path laser and micrometeorological systems were used for the first time at Yongji to measure NH3 losses from irrigated maize. Sprinkler irrigation improved water use efficiency and substantially reduced NH3 volatilisation.
5. An economic sub-model is being constructed and linked to WNMM, using data gathered from a previous ACIAR project (LWR1/1996/164) and new survey data from Fengqu county (Henan province). The combined model will be used to assess the trade-off between environmental and economic objectives, and to assess policy options for water and fertiliser management.
6. Three county-wide suveys in Yuci, Yongji and Hongtong, Shanxi Province have been completed. Soil, landuse and village maps have been digitised and most attribute database sets have been compiled. This information is needed for developing the county-scale WNMM model and decision support system. Landsat TM5 images for 2006 have been purchased to estimate crop biomass, leaf area index and N uptake to calibrate WNMM.
7. The Chinese Academy of Sciences has approved A$500,000 co-funding to establish a state-of-the-art auto-chamber system for N2O and NOx measurements at the Yongji site. The system will be ready in October 2007 for the start of the wheat season.
8. The University of Wollongong is building a more portable open path FTIR system, for simultaneously measuring NH3, N2O, CO2, CH4 and CO. The system will be ready to be deployed in the 2008 wheat and maize seasons at Yongji, as well as in wheat, pasture and sugarcane sites in Australia.
9. The Shanxi Agricultural Comprehensive Development Office has funded A$50,000 to assist the social and soil survey in 2007, and has agreed to further funding to support establishment of demonstration farms in Shanxi.

In Australia:
This project is complimentary to work funded through the CRC-Greenhouse Accounting, Australian Greenhouse Office (AGO) and GRDC, mainly focusing on model development, measurement methodology and fundamental N process studies.
10. The WNMM has been significantly modified for Australian conditions:
- The web-based version of WNMM has been completed.
- The phosphorus sub-routine for WNMM has been developed and tested using the site dataset from Yuci County and will be tested for irrigated pastures in Australia.
11. Applications of WNMM in Australia assisted by this ACIAR project include simulation of water and N dynamics, and N2O emissions, for rain-fed wheat in Victoria and Western Australia and irrigated pastures in Victoria for AGO programs, and simulation of N dynamics in NSW within the ACIAR project led by Dr Jeff Evans. Other international applications, in association with the ACIAR project, include simulating N dynamics and NH3 volatilisation in rice for South Korea; simulating water and N dynamics for irrigated maize and wheat in the Yaqui valley, Mexico, by Stanford University, and; simulating N and water in legume systems in China by the Chinese Academy of Sciences.
12. Based on the ACIAR project, an application to the DEST China Special Fund, ‘Improving water and agri-environmental sustainability in the Murray-Darling Basin and the North China Plain’ was successful ($108,000). Also an additional $120,000 from AGO to fund two open path lasers for NH3 measurement was awarded.

The objective of the project is to improve management of water and nitrogen (N) fertiliser to increase farm incomes, improve environmental quality and reduce nitrous oxide (N2O) emissions from agriculture. The systems being studied are irrigated maize, wheat and cotton in the western Yellow River basin of northern China, and intensive irrigated pasture and rain-fed wheat in Australia.
In China
Second year field experiments in Yongji and Hongtong have been completed to test the effectiveness of reduced irrigation and N fertiliser regimes, also providing datasets for calibrating the water and nitrogen management model (WNMM).
At Yongji, similar wheat (4.6~4.7 t/ha) and maize (8.0~8.6 t/ha) yields were achieved in the optimised treatments with 16% less (70 kgN/ha) N fertiliser and 26% less (155 mm) irrigation compared to traditional farmer practices. Total plant N recoveries were 22% under traditional practices and 35% in the optimal treatment; an improvement of almost 60%. There was substantial mineral N in soils at harvest under both treatments (about 30% of applied N), indicating potential for further reduction of N application rate.
At Hongtong, yields of wheat (8.9~9.1 t/ha) and maize (8.0~8.2 t/ha) were similar in the two treatments, but farmers traditionally applied more N; 462 and 340 kg N/ha for wheat and maize, compared to 180 kg N/ha in the optimised treatment. Irrigation in the optimised treatment was about 20 mm less than under traditional practices. In contrast, the farmers’ practices of N application and irrigation on cotton were almost as efficient as the optimised practices, suggesting little potential for improvement under cotton.
The fertiliser cost saving ranged from 120 to 160 RMB/ha in Yongji and 640 to 1128 RMB/ha (1A$ to 5RMB) in Hongtong. The lower input of N fertiliser without yield loss, particularly in Hongtong, has had a significant impact on farmer practices already. In this year (2008), many farmers near the experimental site applied similar amounts of N as in the optimised treatment.
Datasets of N2O, NOx, methane (CH4) and CO2 emissions derived from continuous automatic chambers have been obtained by the Chinese Academy of Agricultural Sciences (CAAS) at Yuci. The results indicate a potential to reduce N2O emissions by up to 20% through optimised fertiliser and irrigation practices. The Chinese Academy of Sciences (CAS) at the Yongji site has established automated continuous chambers for wheat, maize and cotton, with $500,000 co-funding from CAS. This system measures N2O, CO2, CH4 and NOx.
Open path lasers were used at Yongji to measure ammonia (NH3) volatilisation in conjunction with inverse atmospheric dispersion modelling (WinTrax). The system measures NH3 fluxes as low as 0.5 kgN/ha/day. The results demonstrated that surface broadcasting urea followed immediately by sprinkler irrigation minimized NH3 volatilisation.
Model and decision support system development
The WNMM has been calibrated for the Yuci site to simulate soil-water, N and N2O emissions. An economic module has been integrated with the bio-physical model, allowing the quantification of the public good costs & benefits of different policies of irrigation and fertiliser use.
Using a combination of survey and remote sensing, the spatial datasets for three counties, Yuci, Hongtong and Yongji, including soil attribute data, landuse, village boundaries, village based management practices and crop yields, have been completed for up-scale modelling from experimental site to whole county. Soil hydraulic properties will be estimated by an inverse modelling approach using satellite imagery in conjunction with pedotransfer functions and a clustering of soil types for manageable parameterization.
The conceptual framework of the decision support system (DSS) has been developed for the three counties, in which farm profit and environmental impacts are incorporated.
In Australia
The WNMM has been modified for Australian conditions to simulate water and N dynamics, and N2O emissions for rain-fed wheat, cotton, irrigated pastures and sugarcane. The WNMM has also been modified to simulate N dynamics in wheat-legume systems for the ACIAR project led by Dr Evans in NSW DPI. Working with WA Agriculture and The University of WA, the WNMM has been used to simulate regional water, N dynamics, yield and N2O in the WA wheat belt.
A total of 16 peer reviewed papers have been published. A new open-path (laser and FTIR)-micrometrological techniques has been developed to measure NH3 and N2O emissions at the field scale. Twelve young scientists have been trained to conduct complex laboratory and field experiments, survey and modelling, and 30 people were trained in modelling. The ACIAR project provided leverage to obtain substantial complementary research funding.

The objective of the project is to improve management of water and nitrogen (N) fertiliser to increase farm incomes, improve environmental quality and reduce nitrous oxide (N2O) emissions from agriculture. The systems being studied are irrigated maize, wheat and cotton in the western Yellow River basin of northern China, and intensive irrigated pasture and rain-fed wheat in Australia.
In China - The third year of field experiments in Yongji and Hongtong have been completed to test the effects of reduced N fertiliser application and irrigation on maize and wheat yields, as well as providing datasets for calibrating the water and nitrogen management model (WNMM). At Yongji county - 14% higher wheat yield was achieved in the optimised treatment with 18% less N fertiliser application (40 kgN/ha) comparing to the traditional farmers’ practices. The optimised treatment of split applications of a total of 180 kg N/ha urea resulted in a wheat yield 5.0 t/ha whereas the traditional farmers’ practices of a single basal application of 220 kg N/ha resulted in a wheat yield of 4.3 t/ha. There was substantial mineral N in soils at harvest (about 30% of applied N), indicating potential for further reductions of the N application rate. Similar maize yields (8.9~9.1 t/ha) were achieved in the optimised treatments but with 14% less N fertiliser (30 kgN/ha) than the farmers’ practice. In contrast, and similar to the last year’s results, the farmers’ practices of N application and irrigation on cotton were almost as efficient as the optimised practices, suggesting little potential for improvement under cotton. At Hongtong county - The yields of wheat (8.8~8.5 t/ha) and maize (8.8~9.2 t/ha) were similar in the traditional farmers’ and optimised practices. However, the N application rates in optimised practices,180 kg N/ha, were 47 and to 61% less than that in farmers’ traditional practices, 340 and 462 kg N/ha for maize and wheat, respectively. The fertiliser cost saving ranged from $25 to $30/ha in Yongji and $120 to $211 in Hongtong. The consecutive and consistent two years’ results of lower input of N fertiliser without yield losses (particularly in Hongtong) has had a significant impact on farmers’ practices already. More than 60% farmers near the experimental site applied similar amounts of N as in the optimised treatment developed by our project.
Greenhouse gases (GHG) emissions - For the first time in China emissions of the greenhouse gases of N2O, NOx, methane (CH4) and CO2 were measured continuously by automatic chambers in the ACIAR project by the team of Chinese Academy of Agricultural Sciences (CAAS) at Yuci and team from Chinese Academy of Sciences (CAS) at the Yongji. Both CAS and CAAS provided substantial co-funding ($500,000) to establish such sophisticated systems. The auto chamber system (CAS) is capable of detecting as low as 3-9 g N2O-N/m2/h emission fluxes and 0.6-1.8 g NO-N/m2/h emission fluxes. The results at Yuci indicated the reduction of N2O emissions by 25 to 40% through optimised fertiliser and irrigation practices while there were only small differences in N2O emissions at the Yongji site between the optimised and traditional practices. The N2O emissions accounted for only 1.4 to 2% of applied N, higher than to IPCC default value of 1% but much high than Australian wheat system of less than 0.2%.
Pathways of N losses - Open path lasers were used at Yongji to measure ammonia (NH3) volatilisation in conjunction with inverse atmospheric dispersion modelling (WinTrax). The system measures NH3 fluxes as low as 0.5 kgN/ha/day. The results demonstrated that surface broadcasting urea followed immediately by sprinkler irrigation minimized NH3 volatilisation. Incorporation of urea into soil (point application and covered up) also reduced the NH3 losses but still accounted for 5-10% of applied N. Significantly more NH3 was lost even if irrigation was delayed just one day after fertiliser application. Denitrification was measured by the acetylene inhibition-intact core technique, the denitrification rates were low, ranging 2.9 to 115.9 g N /ha/day, and mostly in top 20 cm soil, and with total losses are less than 5% of applied N. The main pathway of fertiliser N losses in this low organic and sandy soil is nitrate leaching, which was reflected by the high nitrate concentration in the shallow ground water, up to 55 mgN/litre - more than five times the drinking water standard.
Model and decision support system development - WNMM has been calibrated for the Yuci, Hongtong and Alxa sites to simulate soil-water, N and N2O emissions. An economic module has been integrated with the bio-physical model, allowing the quantification of the public good costs and benefits of different policies of irrigation and fertiliser use. Using a combination of survey and remote sensing, the spatial datasets for three counties, Yuci, Hongtong and Yongji, including soil attribute data, landuse, village boundaries, village based management practices and crop yields, have been completed for up-scaling the modelling from experimental site to whole county. Soil hydraulic properties has been estimated by an inverse modelling approach using satellite imagery in conjunction with pedotransfer functions, and a clustering of soil types for manageable parameterization. A GIS decision support system for Hongtong County has been developed using WNMM model to simulated large number of management scenarios, in combination with county scale survey and remote sensing. The DSS incorporates economical, environmental and climate variability.
Sustainability of current nutrients managements and farmers’ perception on environment - The large and comprehensive soil and social-economical surveys to provide essential data for model input and validation at county scale were conducted for all three counties. A survey of 801 households in 10 townships of Yuci County showed: (I) high plant available N and phosphorus (P) in soils due to excessive use N and P fertilizers use, but severe deficiency in potassium (K). This indicates the current nutrient management practices are not only unsustainable but also contribute to the lower efficiencies of use of N and P fertilisers. A balanced nutrient management approach is required. Similar conclusions were found in the Hongtong county’s scale survey. (II) An evaluation of farmers’ economic and environmental perceptions found that the objective of profit maximisation was the crucial factor in fertiliser decision making. The amount of N fertilisers use was negatively related to farmers’ environmental awareness which is significantly affected by the education level of household head and off-farm income levels. Approaches that promote farmers’ education and improve extension services, and the implementation of a tax on excessive use of N fertilizer, or financial incentives to use less, are potentially the best policy interventions for sustainable fertiliser management. There is a large potential for a significant improvement of water use efficiency. However, the improvement is subject to current water institutions and allocation and pricing policies, and irrigation methods. Further investigations on how to effectively extend current research results, including irrigation, to the broader community and different regions are warranted.
In Australia - The laboratory studies to understand how soil and environmental variables quantitatively influence N2O emissions and the proportion of N2O in emissions from nitrification processes have been completed. These are the critical knowledge gaps and essential for process-based N2O model development. The combination of laboratory incubation and modelling with continuous measurement data in the field confirmed that the N2O emission factor for rain-fed wheat system is less 0.2% of applied N, far less that 1% default value set by IPCC. The majority of the N2O is derived from the nitrification process. It is notable that one week peak N2O fluxes after rainfall or fertiliser applications account for ~20% of total annual emissions. WNMM has been modified for Australian conditions to simulate water and N dynamics, and N2O emissions for rain-fed wheat, irrigated cotton and pastures, and sugarcane. Working with WA Agriculture and The University of WA, WNMM has been used to simulate regional water, N dynamics, yield and N2O in the WA wheatbelt. A unique technique to estimate the nitrogen status of crops with a digital camera has been developed. Using a modified digital camera, the framework of a handheld computer based decision support system has been developed, which has been successfully tested in China. Enhanced efficiency fertilisers (EEF), including controlled release, and fertiliser mixed with nitrification and urease inhibitors, have proved effective in improving the efficiencies N fertilisers by reducing NH3 volatilisation and N2O emissions, and potentially nitrate leaching and denitrification. The nitrification inhibitors have also been tested in field experiments in China at the Yongji site. However, the impact of EEF is complex, interacting with soil and plants and other interaction processes. The team is working to incorporate the effect of EEF into the process model and DSS.
A total of 15 peer reviewed papers have been published. A new open-path (laser and FTIR)-micrometrological techniques has been developed to measure NH3 and N2O emissions at the field scale and in feedlots. 11 young scientists have been trained to conduct complex laboratory and field experiments, survey and modelling, and 6 people were trained in modelling. The ACIAR project provided leverage to obtain substantial complementary research funding.

Project ID
LWR/2003/039
Project Country
Inactive project countries
Commissioned Organisation
University of Melbourne, Australia
Project Leader
Dr Deli Chen
Email
delichen@unimelb.edu.au
Phone
03 8344 8148
Fax
03 8344 4665
Collaborating Institutions
Shanxi Academy of Agricultural Sciences, China
Chinese Academy of Sciences, China
Cardno Acil Pty Ltd., China Office, Australia
Chinese Academy of Agricultural Sciences, China
China Agricultural University, China
Project Budget
$1,627,826.00
Start Date
01/04/2005
Finish Date
31/03/2009
Extension Start Date
01/04/2009
Extension Finish Date
30/06/2010
ACIAR Research Program Manager
Dr Mirko Stauffacher