Work Packages:

  1. Systems modelling to develop closed-loop N cycling strategies for Chinese agro-ecosystems
  2. Enhancing recovery of applied N, both fertiliser and manure
  3. Reducing GHG emissions due to N applications
  4. Maximising legume N fixation and uptake
  5. Reducing crop N demand
    1. Predicitng canopy N demand
    2. Maximising C & N fixation and harvest
    3. Minimising grain N demand
  6. Reducing end-user demand for N and N excretion
  7. Optimising land management strategies
  8. Dissemination

Closed loop N cycling

WP1: Systems modelling to develop closed-loop N cycling strategies for Chinese agro-ecosystems.

WP1 will integrate the findings from WPs 2-7 into a single systems model based on the Cool Farm Tool (CFT; Hiller et al., 2011, 2012), which has previously been translated into Mandarin and used in China under a Netwon Agri-Tech project. WP1 will entail the use of process-based models to explore the new data collected in each of WPs 2-7. This will entail model testing, model development, parameterisation and calibration, retesting and sensitivity / uncertainty analysis. When acceptable process-model performance has been obtained, simpler relationships between inputs (environmental and management drivers) and outputs [NUE, air quality (NH3, NOx), water quality (DON, NO3-, NH4+) and climate forcing N2O/N2, NOx)] will be extracted from the process-models, to be used in the systems modelling framework. This use of detailed models, with translation into simpler relationships for use in a systems modelling framework allows the best scientific understanding to be represented in an easy-to-use tool that does not have the input data and computational requirements of the individual process-based models.

The main process model to be used in WP1 is ECOSSE (Smith et al., 2010a,b; Bell et al., 2012) but will be informed by other process models where appropriate. ECOSSE will incorporate data and other model outputs, so that it acts as a single process-model to integrate the outcomes from the models used in each WP. This will allow a single evaluation, and sensitivity and uncertainty analysis, to be conducted and will allow a single framework to be used to derive the simpler relationships (between inputs and outputs) for use in a systems modelling framework (based on the CFT). The CFT will then be integrated with other decision support tools for use in WP8. The use of data and other models in this process is described in more detail below.

1. Process - model testing: The ECOSSE model will be used to integrate the understanding gained from the modelling and experimental work undertaken in WP2-7. WP1 will mainly use field data from the many field experiments contributing to the VJC. From WP2 this includes field data on N recovery N2O and NH3 emissions, from WP3 also emission data, from WP4, various data on components of N fixation, and from the crop field experiments in WP5. Spatial modelling outputs from WP7 will be used to test process description.

2.Process model improvement, parameterisation and calibration: The 13C and 15N, N immobilisation, nitrification and denitrification rate data from WP2 will be used to calibrate and parameterise the N routines of ECOSSE for Chinese conditions, and data on the microbial role in nitrite accumulation from WP3 will be used to calibrate process rate modifiers. Data from WP4 on legume N fixation rates and Rhizobia data will be used to parameterise the model for N fixation in China. Data on canopy N demand data, canopy N model processes, grain allocation of N, 13C and 15N data, grain protein and mRNA data from WP5 will be used to calibrate the plant module and N uptake / allocation routines.

3.Process model (independent) testing: The improved, calibrated and parameterised model will then be retested against similar data to that used in initial testing, but data from different fields, years and manipulation experiments will be used; independent of that used in the first testing and model development.

4.Process model sensitivity and uncertainty analysis: To inform the uncertainty of the model when applied spatially in WP7, a full sensitivity and uncertainty analysis will be performed (e.g. Hastings et al., 2010). To test the model for interventions and changed practices, data for this will come from WP2 (slurry, urease and nitrification inhibitor), WP3 (fertiliser formulation data and mitigation strategies), WP4 (data on cash crops, companion crops and catch crops), WP5 (canopy N and N fixation processes), WP6 (N excretion rates) and WP7 (spatial variation in N fluxes).

5. Input to the Integrated Systems Model and Decision Support Tool: Using the CFT as a basis, we will further develop this into a tool for use by Chinese farmers and extension workers, to examine the impacts of management on N efficiency and N losses. As well as using relationships developed from the process-models as described above, data from the WPs will also be used, including management practice and mitigation data from WP2, 3 and 4, new legume management practices (WP4), canopy management practices, non-legume N fixation and Gd field tests (WP5), information of demand side measures & N excretion (WP6), and will be informed by spatial outputs from WP7 and from WP3. The contribution of information from each step of model / decision support tool development is summarised in the table below.

WPTestingDevelopment, parameterisation & calibrationRe-testingSensitivity & Uncertainty analysisDecision support tool
2 Field N recovery, N2O & NH3 emissions 13C and 15N, N immobilisation, nitrification, & denitrification rates Field N recovery, N2O & NH3 emissions Slurry, urease & nitrification inhibitor data Management practice data
3 Field N2O emissions Microbial role in nitrite accumulation; Field N2O emissions Fertiliser formulation data; Mitigation strategies Management practice & mitigation strategy data
4 Field N fixation data Legume N fixation rates / Rhizobia data Field N fixation data Cash, companion & catch crop data Management practice data; new legume systems
5 Field experiments Canopy N demand data, Canopy N model processes; grain allocation of N, 13C and 15N data, grain protein and mRNA data Field experiments Canopy N and N fixation processes Canopy management practices, non-legume N fixation, Gd field tests
6       N excretion rates Demand side measures & N excretion
7 Spatial modelling of N fluxes   Spatial modelling of N fluxes Spatial modelling of N fluxes Combine with catchment model outputs

WP2: Enhancing recovery of applied N, both fertilizer and manure

Previous research by UK partners has shown that poor recovery of applied N in wheat production systems is associated with temporary net immobilisation by soil microbes. The low C calcareous soils of the North China Plain, by contrast, have a low potential for immobilisation; high rates of nitrate leaching and N2O emissions are associated with excessive net mineralisation and nitrification. There is potential for the soil C:N ratio to be manipulated to enhance retention of N, and thereby lower losses. This WP will deliver an improved understanding of N transformations and their response to C additions in the contrasting soils of China and the UK so that novel techniques to temporarily manipulate N immobilisation, nitrification, and denitrification rates can be designed to enhance recovery of applied N and can be tailored to particular soil and cropping systems, and crop germplasm. Expertise and experimental protocols will be shared between China and UK partners enabling innovations to be compared across climates, soils and cropping systems. Activities are:

1. Quantify N transformations. (i) Develop common protocols for using dual 13C and 15N labelling to quantify C flows (rhizodeposits, soil organic matter, manures), N mineralisation-immobilisation turnover (N-MIT), crop uptake and losses in laboratory experiments (Univ Aberdeen, SRUC, CAU). (ii) Determine the influence of crop species (rice, wheat, maize) on belowground N and C transformations in defined soil types (Univ Aberdeen) and investigate how soil N form affects crop N uptake and growth for a range of adapted germplasm from China (NNU). (iii) Determine how the N form applied (urea, ammonium, nitrate and whole and separated manures) and inhibitors influence the strength of coupling between C-flows, belowground biomass, net immobilisation, nitrification and denitrification by the microbial community (Univ Aberdeen, CAU).

2. Develop novel techniques for modifying plant N availability by

a.) Modifying top soil rooting and rhizodeposition of C (SRUC).(i) Use split-root techniques to vary root distribution between top-soil and sub-soil and dual 13C and 15N labelling to investigate scope for modifying net immobilisation potential by altering top soil rooting. (ii) Use the 3D root architecture model SPACSYS to determine whether the scale of change required in top soil rooting can be achieved through phenotypic selection from existing variation in key architectural traits. (iii) Conduct field experiments in mini-plots with 15N fertilizer to determine whether scale of change required in top-soil rooting and N transformations can be achieved by varying plant spacing in combination with fertiliser placement.

b.) Utilizing non root-derived C amendments to modify immobilisation (CAU). (i) Quantify the effects of non-root C amendments of defined C:N ratio on N-MIT and (for WP3) emission of gases (N2O, NO, N2) using 15N tracing techniques.

c.) Test innovations in the field (i) Propose innovative strategies for enhancing recovery of applied N based on improved mechanistic understanding above, incorporating concepts of plant spacing, precision N application, soil C amendments and inhibitors. (ii) Measure effects of fertilizer and manure application techniques (e.g. broadcast, bandspread, incorporation, with or without urease or nitrification inhibitors) on N recovery, N2O and NH3 emissions (ADAS & CAU) (iii) Optimise application timing of whole and separated livestock slurries and combinations of inorganic N fertiliser and manures (ADAS). (iv) Support evaluation of impact of innovations by measurements of spatial and temporal variability of N availability, N loss pathways and NUE in the main Chinese cropping systems (CAU).

WP3: Reducing greenhouse gas emissions due to N applications

Whilst it is well recognised that N applications to land are associated with significant direct N2O emissions, the role of soil nitrite is not understood, and existing models have insufficient accuracy to meet the urgent need to identify practical strategies that minimise N2O without diminishing important economic crop responses to added N. This WP will be closely coordinated with WP2, will use additional measurements, experiments in China (NNU, CAU), and intensive new datasets from the UK and China to build a parsimonious model (SRUC, ADAS) that specifically and robustly interrelates factors affecting soil mineral N dynamics (WP2), N2O emissions and crop N uptake from arable land shortly after N application, so to quantify and highlight for each environment optimal fertiliser and manure application strategies for GHG mitigation, recognising effects of N on crop productivity.

1. How nitrite accumulation relates to N2O emissions in calcareous soils: (NNU, CAU) Static incubations with a continuous robotic GC measurement system (Robot), 15N labelling, molecular microbiological techniques and field observations will be employed on calcareous soils in China to interrelate: (i) N2O production; (ii) nitrite removal and accumulation; (iii) soil microbes engaged in nitrite accumulation; and (iv) spatial and temporal variations in nitrite accumulation, as related to N2O emissions due to farm practices.

2. Collate & Analyse Data: (SRUC, ADAS, NNU, CAU) We will exploit recent and historical intensive and comprehensive UK and Chinese records of daily N2O fluxes and associated environmental measurements (topsoil moisture, topsoil SMN, temperature & rainfall) particularly relating to times and rates of fertiliser N applications to generate a parsimonious data-driven model of the interacting and controlling factors of N2O loss and the effect of N fertiliser (ammonium nitrate and urea) application timing. Using associated crop yield data, we will express N2O emissions in relation to crop yield (i.e. emissions intensity; GHGi) and use this to deduce approaches that reduce GHGi of crop products. Model robustness will be enhanced through the wide ranging conditions and management practices represented by the datasets e.g. N type, rainfall characteristics, temperature, cropping practices, soil types.

3. Validate, and up-scale for rotations & climates: (a) (NNU, CAU, SRUC, ADAS) Validate the empirical model using independent UK and Chinese data and explore sensitivities of counter-measures for mitigating N2O (and other GHG) emissions. (b) (NNU, CAU) Comprehensive activity data will be collated on crop production and the model will be used to quantify overall GHG emissions and GHGi under the different common crop rotations in the North China plain as affected by: (i) manure use; (ii) the multiple cropping index, especially (iii) vegetable fields with a range of cropping frequencies. (c) (NNU, CAU) Use the model to improve understanding of the impacts of future climatic conditions (30+ years) on N2O emissions and GHGi in response to fertiliser N.

4. Design mitigation strategies: (NNU, CAU) With policy-makers and stakeholders, develop guidance on advice to minimise direct N2O emissions from N fertilizer use. Test simple guidance with leading farmers (WP8) on best form, timing, method and amount of applied N (organic, inorganic, use of inhibitors, etc.), addressing socio-economic factors e.g. part-time farmers with 2nd employment hence restricted flexibility for sophistication of farming activities.

WP4: Maximising Leguminous N Fixation and Uptake

Major legumes are of different species and currently fill contrasting roles in arable rotations of the UK and China, so coordinated bilateral research will offer exceptional opportunities to gain insights into species choices and management practices that maximise N fixation and crop N uptake. Work here will explore (in parallel) hypotheses that amounts of N fixed by leguminous Rhizobia (assuming similar nodulation) relate to the availability of photosynthates within the plant (as affected by light intensity and alternative sinks), and amounts transferred to non-legume crops (ignoring soil N loss processes) relate inversely to the strength of protein sinks within the plant.

1. How to maximise fixation & transfer (Lab.): (SRUC & CAAS, Year 1) (a) Study legume N content and transfer to companion plants using a 15N approach; Legume (soya) and wheat plants will be grown separately and in a mixture under controlled environmental conditions. Rates of N fixation and transfer of N will be determined using a natural abundance technique. This experimental system will be used to evaluate variations in crop photosynthesis (with light intensity) and soil N & C status (to affect nodulation or activity) to test conditions for maximum N transfer. (b) Prepare 15N enriched legume residues (leaf, stem, roots, nodules) and test effects of residue composition, soil moisture and non-legume presence / absence on legume N transfer to the companion or following crop.

2. How to maximise fixation & transfer (Field): (CAAS, Years 2&3). (a) Determine amount and major limiting factors for N fixation of major legume species used in typical legume-based cropping systems in dominant agroecoregions. (b) field-test best N transfer approaches (using 15N label) as predicted from lab for leguminous (i) cash crops e.g. pulses, (ii) companion crops & (iii) catch crops. (b) In parallel with WP4-1b use 15N enriched legume residues to test effects of crop and soil management on legume N transfer to the companion or next crop. Coordinate measurements with work on soil microbial processes in WP2.

3. Design legume cropping systems: (CAAS; Years 2&3) Develop integrated cropping systems based on best practices for maximising N fixation and transfer from WP4-1 & -2. Undertake field experiments comparing legume systems and management practices to assess effects on leguminous N fixation, N availability and N transfers to non-legume crops. Manipulations would include high stubble (of wheat in Northwest China and rice in South China), mixed sowings of leguminous and non-leguminous green manures, and combinations of crops with distinct differences in root and architecture.

4. Evaluate new legume systems: (CAAS; Year 3). Initiate long-term evaluations (including yield, fertilizer replacement, soil fertility and microbial diversity effects) of new legume-based cropping systems in association with the long-term N fertilizer replacement experiments already established by CAAS over 6 to 8 years. Complete an economic evaluation (comparing the relative value and cost of legumes as companion or catch crops with bare fallow).

WP5: Reducing Crop Demand for N

New physiological and genetical understanding generated by this WP will support Chinese (& UK) growers’ judgements of their crops’ demands for fertiliser N (WP5A&B) and will provide means to reduce those demands (WP5C). It will build a fundamental understanding of crop N responses to (i) persuade and empower farms in China to regulate N use according to crop N requirements, (ii) show agronomic routes to reduce N requirements, (iii) allow the design of ‘HYLO’ ideotypes (High Yield Low Optima) and (iv) determine the importance of HYLO sub-traits as selection criteria for plant breeding. The complementary expertise of the University of Cambridge and ADAS will be employed to model existing UK and Chinese data, to generate hypotheses and test them in laboratory and field experiments in the UK for later use in China (by CAS).

WP5A: Predicting Canopy N Demand

The Canopy Management approach to crop N optimisation has been widely adopted in the UK, Australia and New Zealand; extensive datasets and enhanced computing power now offer an opportunity to underpin Canopy Management with accurate, robust and generic parameterisation, and scientific publication, to support its extended adoption in China and beyond, and to apply its principles in ideotype design and genetic improvement. Real time optimisation of crop canopies is key to optimisation of crop N nutrition at all scales of farming.

1. Collate existing unpublished data & approaches from the UK (ADAS) and China (CAS) relating to applied N effects on wheat canopy expansion through time. (i, ADAS, CAS) conduct a bilateral workshop to shareCanopy Management principles, (ii, CAS, ADAS) Collate validation wheat, rice & maize data from China to augment the unique GNR wheat dataset providing highly-resolved (weekly through four successive seasons at 6-10 N levels from nil to double-optimal) data for available N, canopy N, green area index (GAI), wheat biomass and biomass partitioning to grain, complemented by in situ high-frequency meteorological data including soil moisture. (iii, ADAS, CAS) compare procedures for analysis and interpretation of N response curves, and modify the Optiplot experimental technique for variety testing in China. (iv; Camb. Univ, CAS, ADAS) Propose a framework for analysis of crop N responses through time for wheat, rice and maize.

2. Parameterise canopy dynamics and crop N responses: (Camb.Univ.) (i) Use state-of-the-art methods of iterative statistical fitting of intensive crop growth data, including Bayesian and statistical inference, to model the dynamics of canopy expansion, survival and senescence, to develop a parsimonious model (sets of coupled differential equations) of canopy expansion and crop N uptake in relation to N supply and temperature. (ii) Extend the model for biomass growth, dry matter partitioning and final yield. (iii) Test the model with wheat, rice & maize data from China to define the extent of it validity.

3. Explore model behaviour (Camb.Univ.) Use sensitivity analysis, and model reduction techniques to define (i) canopies that deliver economically optimal crop performance (hence to develop benchmark algorithms for canopy sensing), (ii) most telling traits for genetical analysis and (iii) likely impacts of genetic or agronomic modification of canopy sub-traits, for example reducing specific leaf N. Such a model would also enable recognition of optimum N supply for canopy growth at any time.

4. Implement Canopy Management (ADAS, CAS) by linking the model with extant canopy sensing methods in management Apps and other manual assessment tools, to better predict N requirements in season.

WP5B: Maximising C & N fixation, and distribution to grain

So as to define traits for selection and breeding of both wheat and rice that enhance yields and reduce N use, the goal here will be to test the hypotheses that (i) HYLO phenotypes excel in uptake, assimilation or partitioning of C due to canopy structure or protein sink strength, and (ii) that crop N demands could be reduced by enhancing the potential for N inputs from associative N fixation processes of endophytic diazotroph (Gd). Tasks:

1. Maximising photosynthetic C and N conversion and partitioning. HYLO ideotypes will be compared with contrasting low and high Harvest Index genotypes under controlled environment growth conditions. We will explore the uptake and transfer via 13C (as CO2) and 15N (soil fertiliser) labelling between major canopy leaves and developing ears / panicles. We will identify C and N losses during storage and remobilisation of assimilates, and particularly the roles of respiration and photorespiration during grain filling phases. Measurements of CO2 and NH3 compensation points will indicate interactions between canopy structure and climatic factors (such as limiting light experienced in the 2012 UK grain-filling period).

2. Testing C and N allocation under field conditions. In the field we (Cambridge, ADAS) will compare the impact of genotype (HYLO and contrasting High Yield High Optima lines), canopy structure (due to stem density and N status) and flowering time on C and N partitioning, and N optima. Experimentation and modelling of optima will predict the canopy management strategies (involving timing and rate of fertiliser applications) and genotypes which minimise N inputs, to support more widespread selection and adoption of HYLO lines.

3. Maximising non-leguminous N fixation (Griffiths, Azotic, ADAS) N labelling (as 15N2) on lines being tested in Tasks 1 and 2 will reveal amounts and impacts of N availability from Gd, whilst 13C labelling will show the interplay with root carbohydrate exudates and tissue oxygen tension. This work would utilise the Cambridge facilities and expertise for stable isotope analysis. In-kind contributions from Azotic will identify intra-plant and temporal patterns of crop colonisation with Gd, as affected by canopy characteristics.

4. Quantifying N contributions from Gd – field testing: (ADAS, CAS) In parallel with lab. comparisons made in Task 3 above, field testing (using the ADAS Optiplot technique in the UK, and after exchange and training, a manual protocol in China) will field-test amounts and impacts of N availability from Gd and best approaches for use of Gd.

WP5C: Minimising crop N demand by grain protein manipulation

This WP tests the hypothesis that manipulating grain protein demand can reduce crop N requirements. High yielding UK wheat crops now need more N for protein in the grain than is required to generate green tissue for photosynthesis (Pask et al., 2012). By selectively reducing prolamin storage protein accumulation in the grain N requirements could be reduced and, due to their poor nutritional quality, the balance of essential amino acids could be improved (Kindred et al., 2007), allergenicity reduced and ultimately livestock N excretion reduced (WP6). The approach will use artificial microRNA (amiRNA) transgenes in preference to other gene knock down or knock out approaches (random mutation, CRISPR etc.) because (i) any construct can be targeted to one or more members of a gene family and a single manipulation can achieve reduced activity in families of functionally redundant genes; (ii) there is potential to generate quantitative variation in a trait with a single construct through between-line variation where some lines will likely have strong knock down of the target genes whereas others will have a milder effect; (iii) the knock down effect is genetically dominantso it will be relatively easy to generate lines in which traits from different transgenes can be recombined in F1 hybrids, allowing testing of many different levels and types of gene knock down in sets of F1 hybrids. Tasks are:

1. Down regulate genes affecting protein deposition (Cambridge Univ.) using artificial miRNAs. The targets include proteinase inhibitor genes and amino acid / peptide transporters affecting mobilization of leaf protein N [1] and transcription factor regulators of prolamin and / or glutelin [2]. Transformation of wheat will be done at the NIAB transformation facility and molecular characterization in the Dept. Plant Sciences.

2. Test fertiliser N responses (ADAS & China) and requirements of a range of wheat varieties differing in grain protein content using Optiplots, with a range of crop sensing measurements.

3. Physiological analyse canopy N (ADAS), grain protein and grain yield relationships (including N harvest index) from selected lines from [2] to test traits and mechanisms for selection of HYLO types.

WP6: Reducing end-user demand for N and N excretion

End-User demands for food, animal feed and biofuel are the ultimate driving force for N inputs to the agricultural N cycle. The hypothesis addressed here is that reducing these N ‘demands’ would significantly reduce N inputs and losses. Since N use efficiencies in Chinese crop and livestock systems are about half of those in Western Europe, current practices use more N inputs to produce the same amount of food, hence causing larger losses of N. Reduction of N demands for human consumption and livestock feed are key challenges for closing the N cycle. Solutions will entail increasing NUE of food production (WP2-5), but also policies and other socioeconomic tools to reduce the overall needs for N. In this WP Chinese participants will focus on a socioeconomic analysis of how N demand might be reduced, while UK participants will review worldwide technologies for applicability in China, and test the key topic of protein in stock feeds.

1. Review grower perceptions: (ZU) Through surveys with stratified samples of farmers and their key influencers, review (i) perceptions of how N fertilizer use affects harvested crop quality, (ii) how crop quality is accommodated in livestock (pig) feed formulation and for direct food and biofuel processing, (iii) factors governing livestock’s needs for protein, (iv) factors affecting actual feeding of protein to pigs, and (iii) the socioeconomic barriers that inhibit the implementation of new technologies in the crop-to-livestock supply chain in China.

2. Review end-user influences: (ZU) Analyse how human diet differences and changes affect N use in crop production via protein use in pig production. Consider the complexity of N cascades between human food consumption and agricultural production. Analyse how changes in end use might affect N cascades, through survey and modelling. Conduct cost-benefit analyses of potential interventions, including the human health cost and benefit from the dietary changes.

3. Review influences on N excretion: (ZU, SRUC) (i) Review Chinese and other literature relating N excretion to N in pig feedstuffs, and (ii) estimate the potential impact of novel feed formulation strategies on pig N excretion in China.

4. Test scope to reduce livestock N excretion: (SRUC, ZU) Finisher pig diets with either standard or reduced levels of protein, with essential amino acid levels maintained, will be formulated using typical UK and Chinese feedstuffs. Feed matrices for the four resulting treatments will be informed by scoping available feedstuffs in the UK and from Tasks 6.1 and 6.3 for China. These four treatments will be applied to pigs in both UK and China (n=9 in each country). Detailed measurements of N intake, N retention (through deuterium dilution technique derived body composition), and faecal N excretion (through marker assisted total tract N digestibility), will inform on quantity and route of N excretion.

5. Policy formulation: (SRUC, ZU) Use outcomes of Tasks 6.3 and 6.4 to formulate advice to both China and UK pig production industries on practical ways to reduce N excretion through feed formulation.

WP7: Optimising land management strategies at Field, Farm and Catchment Scales

This WP will use the integrated modelling framework developed in WP1 and catchment scale models, to raise the outputs of plot/farm level models to catchment scale. The N pollution impacts and understanding from plot, field and farm scale will be up-scaled to catchment scale using a spatially distributed hydrological model, so that policy options for land management at catchment scale can be explored. Data and expertise in plot, field, farm and catchment modelling will be shared between China (CAAS, BNU, CNU) and the UK (Aberdeen, SRUC, ADAS). Tasks are as follows:

1. Monitor characteristics of N losses at farm to catchment scales. Activities will be to (i) monitor rainfall-runoff and N exports from representative fields and farms, and assemble existing data from China & the UK to identify the characteristics and patterns of N loss at farm scale – this will include losses to water (DON, NH4+, NO3-), the atmosphere (N2, N2O, NOx, NH3) and through export in products; (ii) evaluate influential factors such as climate (using a rainfall-simulation experiment; Fig. 2), crop growth, hydrology and management, both empirically and using the process-models (ECOSSE) and system level models (CFT) used in WP1; (iii) identify key management interventions reducing N loss at farm scale using outputs from WP2-6 and the models from WP1, (iv) test impacts of different environmental events on N loss at catchment scale using high resolution monitoring data, (v) assess lag effects, N retention and dilution, and how spatial heterogeneity of rainfall affects watershed N loss dynamics.

2. Modelling N losses at catchment scale.(i) Integrate watershed, groundwater and water quality models (SWAT-MODFLOW-EFDC) and the system level N cycle models from WP1 (ECOSSE, CFT) within a geographic information system (GIS), building on existing research in China; (ii) Identify the most sensitive parameters leading to N loss though sensitivity analysis; (iii) Use existing field measurements to validate the key parameters and to calibrate the models where necessary; (iv) Conduct storm-event and long-term continuous simulations to predict N exports and concentrations at catchment scale and long-term climate data to simulate gaseous losses, (v) Quantify impacts on N losses to water courses and to the atmosphere at multiple spatial scales.

3. Strategic optimisation of N management. (i) Use the spatially distributed hydrological and system level models to test potential N cycle management interventions on productivity & N sustainability indices (including air quality (NH3, NOx), climate forcing (N2O/N2/NOx) and water quality (NO3-, NH4+) at each spatial scale; (ii) Propose cropping, management and N use scenarios / interventions that will be beneficial at all scales, especially with respect to fertilizer N and manures; (iii) Determine N inputs that optimise productivity vs N impact at different spatial scales, including region, (iv) Identify critical management zones contributing disproportionately to N pollution of air and water. (v) Propose an environment risk zoning method using Bayesian statistics; (vi) Evaluate and describe strategies that best reconcile agricultural productivity and N pollution at multiple spatial scales.

WP8: Dissemination

The strategy of this WP is to establish relationships, organisations, networks and a quality mark, with governance that ensures the VJC’s communication and engagement systems persist beyond the life of the project, becoming stakeholder-driven, principled, branded and permanent. Through three years the VJC will seek engagement with existing scientific institutions and working groups, government departments, industry organisations and farming associations, etc. so that its vision, innovations and skills achieve maximum long-term impact. A multi-targeted approach will be adopted addressing the needs of science (including students), policy makers at national and regional levels, extension and industry workers (especially agents for farm supplies), farmers and consumers.

1. Establishing demonstration areas. 2 or 3 demonstration sites will be selected in the three regions where CAAS & CAS have established emission monitoring networks (Huang-Huai-Hai region, Southwest region, and the middle-lower Yangtze River region).

2. Technology integration and evaluation of applicability. New techniques, products or equipment from each WP above will be evaluated for use in four regions. Possible new techniques may include a) fertilization models: chemical fertilizer or organic fertilizer combinations and rates, fertilization timings and placing (WP2), in-field water and fertilizer optimisation (from WP5), b) planting patterns: intercropping, rotation, catch crop, green manure (from WP2, WP4, WP5) c) new-type fertilizer and manure products, perhaps with nitrification inhibitors (from WP3). New equipment may include variable rate fertilizer applicators and fertilizer deep placement with the machinery, suitable for paddy fields, dryland or protected vegetable land (from WP2). All advocated methods will be supported by cost-benefit analyses especially of their environmental effects (from WP1, WP6, WP7).

3. Scientific Promotion & Training. (i, CAAS, Univ Aberdeen, Year 3) Organise and hold a conference in China to report N-Circle results and associated scientific developments during the life of the project. (ii) Organise and hold Workshops for national and regional policy-makers on using the N-Circle models and tools (e.g. CFT, MANNER, and models on Canopy Management, feed protein, water quality) for both scientific and industry purposes. (iii, UEA) arrange for a special issue of papers on N-Circle research in a scientific journal. (iv, UEA) Use the UK-China Sustainable Agriculture Innovation Network (SAIN) as a platform to engage with and disseminate N-Circle VJC achievements to a wide range of stakeholders.

4. Engaging Extension Services and industry. (i, CAAS) Select advanced farmers in Year 1 (through a competition) to host on-farm demonstration areas. Set up demonstration areas in Year 2, so host farmers can master new technologies and train in demonstration techniques (farmer to farmer). (ii, CAAS, UEA, ADAS) Translate key UK documents e.g. MANNER; formulate and distribute technical brochures and videos. (iii, CAAS) Conduct a programme of training courses and seminars for extension workers and third party influencers. (iv, CAAS) Set up technical service through training third parties (or contractors) to provide timely and efficient fertilizer & manure application to individual farms [overcoming issues of small scale decentralized operations, seasonal labour shortages, farmers’ poor knowledge and remoteness].

5. Developing, supporting and validating an ‘N-Circle’ Quality Mark. (i, CAAS & UEA, Year 1) share the N-Circle vision of closed-loop N cycling with key industry organisations, (ii, Year 2) form working groups to develop criteria that test conformity of any process or product with N-Circle principles. (iii, CAAS Technology Transfer Centre) Devise and test a web-based accreditation process for an N-Circle quality mark. (iv, Year 3) Prepare materials including technical regulations, industry standards, manuals, illustrations, videos and websites, and complete a publicity programme through brochures, newspapers, radio advertising, television, and internet to promote the established N-Circle VJC.

References: Hastings AF et al. 2010. Agric. Ecosys. Env. 136, 97-110; Hillier J et al. 2011. Environ. Model Software 26, 1070-1078; Hillier J et al. 2012. Glob. Change Biol. 18, 1880–1894; Bell MJ et al. 2012. Nutr. Cycl. AgroEcosys. 92, 161-181; Kindred DR et al. (2008). J. Cereal Sci. 48, 46- 57; Norse D & Ju X 2015. Agric. Ecosys. Env. 209, 5-14; Pask AJD et al. 2012. Field Crops Res. 126, 104-118; Shock CC & Shock CB 2012. J. Integrated Agric. 11, 14-30; Smith JU et al. 2010a. Clim. Res. 45, 179-192; Smith JU et al. 2010b. Clim. Res. 45, 193-205; Smith P 2013. Global Food Sec. 2, 18-23.