Introduction

Nutrient management in China attracted global attention from the early 2000s when results were made public about the large increases in food production and the large increases in fertilizer use and nutrient losses (e.g., Zhu and Chen 2002; Liu and Diamond 2005; Gao et al. 2006; Ju et al. 2009). There was admiration and respect for the achievement of producing sufficient amounts of nutritious food for about 20% of the global population on about 8% of the global area of cropland, especially after the famine of the 1960s (e.g., Brown 1995; Smil 1999). At the same time, there were astonishment and concerns about the large nutrient inputs via synthetic fertilizer and animal manure, and the associated large nutrient losses. The concerns relate mainly to the human health and climate change impacts, the eutrophication and pollution of water bodies and to the biodiversity losses, associated with nitrogen (N) losses to air via ammonia (NH3) and nitrous oxide (N2O) emissions, and with N and phosphorus (P) leaching and discharge to water bodies. Concerns relate also to the large resources use (notably of water, leading to groundwater depletion) in intensive agricultural systems, and to the large carbon dioxide (CO2) emissions associated with the production of synthetic fertilizers. The intensive crop and animal production systems, with large N and P inputs and losses, are mainly in the eastern half of the country, more specifically east of the so-called Hu line, which stretches from Heihe in the northeast to Tengchong in the south.

The increasing number of reports about the performance of China’s agriculture were made possible after the introduction of the ‘open-door’ policy from the 1980s (e.g., Zheng and Li, 2022), but it takes time to comprehend what is really going on, also because of the huge diversity in agricultural systems, environmental conditions, culture and approaches in China. Obtaining a comprehensive understanding was also cumbersome because of the often disciplinary and regional scientific approaches, which made it difficult to obtain oversight and to understand the cause-effect mechanisms at different spatial and temporal scales. Twenty years later, the questions still are ‘do we really know and understand quantitatively the nutrient management in China’s agriculture, and what needs be done to drastically increase nutrient use efficiency and decrease nutrient losses?’.

Here we argue that progress has been made in understanding nutrient cycling and management in China’s agriculture, especially by scientists, but that farmers have no ‘steering wheel, dashboard and tools’ for managing the nutrient sources and flows accurately on their farms. Interesting attempts have been made to bridge the gap between science and practice in the Science and Technology Backyards (Zhang et al. 2016), but the large majority of the farms still have no tools, guidelines and incentives for implementing best nutrient management practices. Major improvements have to be made in the agricultural supply and advisory system, as indicated by Ma et al. (2013). We outline some aspects that we consider most relevant for improving nutrient management in practice, and reflect on the relevance of providing a steering wheel, dashboard, tools and incentives to farmers for managing nutrient sources and flows accurately on farms.

Why is nutrient management in China so complicated?

There is often a difference between what is possible according to science and what is done in practice, also in nutrient management (Sect. "Box 1. Nutrient management"). The scientific research of nutrient management in China started to blossom from the 2000s, also following the ‘open-door’ policy. An increasing number of papers are being published in high-impact journals and an increasing number of papers indicate the way forward for practice. The developed crop and nutrient management approaches have been presented as ‘an experiment for the world’ (Zhang et al. 2013), and indicated how crop yield can be increased further and nutrient losses can be decreased at the same time (Chen et al. 2011; 2014). Other papers reported how these approaches can be successfully implemented in practice through empowering smallholder farmers using participatory innovation and technology transfer, and by garnering public and private support (Zhang et al. 2016; Cui et al. 2018).

Box 1. Nutrient management

Nutrient management can be briefly defined as ‘the management of nutrient resources so as to achieve objectives’. Here, nutrient resources commonly refer to the nutrients in soil, crop residues, manure, organic wastes and synthetic fertilizers, while objectives commonly refer to agronomic (yield, farm income) and environmental (nutrient losses to air and water) objectives (Beegle et al. 2000). Evidently, nutrient management may be qualified as good nutrient management or best management practices (BMP’s) when the set objectives are being achieved. In practice, there is often ambiguity in the use of the terms nutrient management and BMP’s, which often follows from unspecific descriptions and interpretations of nutrient resources and objectives. Thus, there are often differences between farmers and society (or governments) in their opinions on which objectives have to be achieved when and how. There are often also differences between farms in the performance of nutrient management, because of differences in farming systems and in environmental conditions.

The actual managers of the nutrient resources and flows are the farmers, gardeners and land managers. These actual managers are, in turn, strongly influenced by the demand from markets, consumers, and from the suppliers of nutrient resources, equipment and information. This can be a broad group of stakeholders, including advisors, farmers’ unions, fertilizer producers and sellers, accounting offices, governmental institutions, research organizations, neighbors and NGOs. The information and incentives provided by the so called ‘agricultural supply and advisory system’ can be conflicting (Ma et al. 2013). One group of stakeholders, i.e., agricultural research organizations and universities, has a special position here, as they may provide new insights in and views about what is possible and needed from agronomic and/or environmental/ecological perspectives. They may also provide integrated assessments and oversight. Hence, the performance of nutrient management can be considered at three levels, (1) practice, i.e., farmers and gardeners, (ii) intermediates, i.e., the agricultural supply and advisory system, and (iii) science, i.e., agricultural research organizations and universities.


However, in-depth reviews of nutrient management in practice led to sobering thoughts (e.g. Gao et al. 2006; Ma et al. 2013), and these thoughts are supported by results of recent farm surveys (e.g., Jin et al. 2021; Zhang et al. 2021a, b; Tan et al. 2021). There is a huge gap between what seems scientifically possible and what is actually occurring in practice. The science of nutrient management is progressing much faster than nutrient management in practice in China, for various reasons. First, the suppliers of nutrient resources, equipment and nutrient management information are important actors in the nutrient management process (see Sect. "Box 1. Nutrient management"). This ‘agricultural supply and advisory system’ has been successful in promoting the use of more productive crop varieties, and subsidized fertilizers and pesticides so as to increase crop yields, but it has not been trained in promoting and implementing environmentally-sound nutrient management practices or Best Management Practices (BPB’s; see Sect. "Box 1. Nutrient management"). This has been ascribed to the fragmented nature of this agricultural supply and advisory system, as there are different stakeholders with different and often conflicting objectives, and to the disconnection between science and this advisory system (Ma et al. 2013). Thus, most farmers have no steering wheel, dashboard, tools, guidelines, and incentives for managing nutrient sources and flows accurately and precisely, and in an environmentally-sound manner. Second, the population of farmers is very large, not well-educated and ageing, while an increasing proportion is part-time farmer with another job in the city (Ren et al. 2023). Thus, communicating with all farmers in an effective manner and convincing them to implement environmentally sound nutrient management is not easy, although basically all farmers have a mobile phone and are connected to the internet. Third, most farms are small (≤ 0.5 ha), and the land tenure system does not invite farmers to invest in improved soil fertility and nutrient management practices (Lyu et al. 2019). Fourth, crop and livestock production systems have increasingly become specialized and spatially disconnected; livestock farms have become large and landless, while crop farms remained very small and scattered, which hampers effective manure recycling and utilization (Jin et al. 2021).

In spite of the aforementioned barriers, total fertilizer input to cropland increased steeply from the 1960s, first for N fertilizers, then for P fertilizers, and thereafter for K fertilizers (Fig. 1). Various explanations have been provided for the exceptionally steep increase and large N application rates, including the availability of cheap fertilizers, changes in crop types and introduction of double and triple cropping systems, with an associated increased application per cropland area, an increase in cropland area, and incentives of the Ministry of Agriculture and the regional agricultural bureaus to produce more food, because of the increasing population, (e.g., Ju et al. 2009; Zhang et al. 2015; Yu et al. 2022). For most crops, N applications per unit cropland area plateaued in about 2010 (Yu et al. 2022).

Fig. 1
figure 1

Changes in the use of synthetic fertilizers in China (left panel), and changes in the total manure excretion by the main livestock species (pigs, cattle, buffalo, poultry, sheep and goat) (right panel) during the period 1960–2020. Amounts of nitrogen (N), phosphorus (P), and potassium (K) are in Tg (1 Tg = 10.12 g = 1 million tonne). Note the different scales of the Y-axes. Data source: FAOSTAT (2023) and Zhang et al. (2021a, b)

Full size image

From about 2016, total fertilizer input started to decrease, for N, P and K (Fig. 1), while food demand continued to increase, and food security remained high on the political agenda. This decrease in fertilizer use has been associated with a decreasing cropland area, changes in crop types (Yu et al. 2022), enlarging farm size (Duan et al. 2021; Li et al. 2023), the release of the governmental action plan ‘Zero Growth in Chemical Fertilizer Use by 2020’ (Jin and Zhou, 2018), increased recycling of animal manure (Zhu et al. 2022), increasing imports of food and feed (e.g., Zhao et al. 2021), and with empowering smallholder farmers (Zhang et al. 2016; Cui et al. 2018). All these factors probably played a role (Yang et al. 2022), but it takes more efforts and time to entangle all cause-effect relationships, while considering all uncertainties in the data. Be that as it may, nutrient losses from crop production likely decrease too, but again further studies are needed to find out where, which losses, and by how much and in which time frame.

While nearly two thirds of the cropland area are used to growth cereals (mainly rice, wheat and maize) and to a lesser extent oil crops (mainly soybean, peanut and rapeseed), nutrient inputs and losses are largest in vegetable and fruit production (e.g., Yu et al. 2022). These are also the most diverse, entrepreneurial and dynamic sectors, but probably also the least-well understood sectors from a nutrient management point of view. Especially vegetable production increased very fast from the 1990s (Fig. 2).

Fig. 2
figure 2

Changes in the total amounts of domestically produced cereals, vegetables, fruit and animal products (sum of meat, milk, egg, expressed in protein) during the period 1961–2020. Amounts are expressed in Tg (1 Tg = 1012 g = 1 million tonne), but cereals, vegetables and fruit in 10 Tg. Data source: FAOSTAT (2023)

Full size image

Another nutrient-dense and highly diverse, entrepreneurial and dynamic sector is the livestock production sector, which has been booming because of the rapidly increasing demands for animal-source food (Bai et al. 2018; Jin et al. 2021; Zhao et al. 2021). It is not easy to quantitatively assess the nutrient management performance of the livestock sector accurately, because of the rapid changes that have taken place and still are taking place, the paucity of accurate data, and the large differences between farms in size, animal productivity, manure management and nutrient use efficiency (Bai et al. 2018; Tan et., 2021). A recent study showed that the livestock number (expressed in pig units) decreased by 14% between 2007 and 2017, while the output of animal produce increased by 3%, indicating increased animal productivity (Zhu et al. 2022). The study was based on surveys of more than 480,000 large farms with livestock, but it was not reported whether the surveys were representative for the livestock sector. Authors also found that manure N recycling increased from about 30% in 2007 to about 40% in 2017, while total NH3 emissions decreased by 15% and N losses from livestock production to surface waters decreased by about 50% between 2007 and 2017 (Zhu et al. 2022). These numbers reflect indeed major changes in livestock production and in the performance of the livestock sector. However, results derived from the China Rural Statistical Yearbook indicated that livestock number (expressed in pig number) has increased by 18% between 2007 and 2017, while the volume of animal produce increased even more (Fig. 2). There has been also a steady increase in the total excretion of N, P and K of about 10% by the main livestock species between 2000 and 2020, following the steep rise (~ 150%) between the 1980s and 2000 (Fig. 1). During the last decade, the total amounts of P and K in manure produced were as large as the total amounts of P and K used in synthetic fertilizers.

Summarizing, nutrient management in China’s agriculture is complicated because of (i) the huge diversity in farming systems and environmental conditions, (ii) the large number of farmers, the small farms and the land tenure system, (iii) the fragmented agricultural supply and advisory system, and (iv) the lack of appropriate techniques and tools for precision nutrient management and tools for monitoring and accounting. There are also uncertainties in some data, both in the data from snap-shot farm surveys and in the Chinese statistical data, especially in the data from before 2000. These uncertainties are related to the diversity in agricultural systems, the large number of farms, the reliance on bottom-up administrative reporting, use of old farm survey data and coefficients, use of nonstandard definitions, the politicization of statistics, and parallel reporting systems in multiple agencies (Liu et al. 2020). Evidently, this complicates nutrient management planning and may lead to confusion. Accurate and up-to-date farm data are essential for good nutrient management and planning, and thus for advisors, research and policy. There is a need for double checking data and for greater verification of data.

What has been the role of China’s government in nutrient management planning?

China’s government is supporting its farmers financially, notably for producing cereals, oil crops and livestock, and the support has been increasing over time. On average about 15% of the agricultural income is derived from governmental support, but for some farming systems the support is much higher (OECD 2022). The government has also supported the production, storage and transport of synthetic fertilizers from about 1950. As a result, crop farmers had access to relatively cheap synthetic fertilizers, which has contributed to the liberal use of synthetic fertilizers and to the neglect of nutrients from manure, crop residues and organic wastes, and thereby to increased nutrient losses (Li et al. 2013).

China’s government did not give high priority to the implementation of environmentally sound nutrient management planning until about 2015. Policy emphasis was on securing food production and raising rural incomes. This changed in 2015, when the subsidies on synthetic fertilizer production ceased and the government issued the action plans ‘Zero Growth in Chemical Fertilizer Use by 2020’ and ‘Promoting the Recycling of Waste from Livestock and Poultry Farming’ (Jin and Zhou, 2018). The zero growth in fertilizer use should result from (i) improved fertilizer types and fertilizer spreading techniques, (ii) increased utilization of livestock manure and crop residues, which should lead to enhanced soil quality, (iii) fertilizer use based on soil quality measurement, and (iv) subsidies for farmers to cover the extra costs for implementing the aforementioned measures. These policy documents reflect increased awareness of the importance of environmental quality and soil quality. Modelling studies suggest indeed that removing fertilizer manufacturing subsidies, enhancing fertilization efficiency, and increasing manure recycling will greatly decrease synthetic fertilizer use and nutrient losses, with substantial benefits for the environment and potentially also for farmers (Van Wesenbeeck et al. 2021; Wang et al. 2023).

The government has also acted as a strong driver for the increased and specialized livestock production in China during the last few decades (Bai et al. 2018). This is rather unique as it is commonly perceived that the increasing demand for animal-source food is the main driver for the so-called livestock revolution in countries with emerging economies (Delgado, et al. 2001). Within 30 years, the number of livestock (expressed in livestock units) in landless industrial farms increased 70 fold, and the proportion of monogastric animals to the total livestock population (expressed in livestock units) increased from 62 to 74% between 1980 and 2010. Three policies greatly contributed to the rapid increase in livestock production and consumption: (i) the liberation of markets and the removal of barriers for animal-source food production and consumption, (ii) economic support policies aimed at boosting crop and animal production around big cities, and (iii) the loose environmental regulations, which did not provide much restrictions to livestock farming practices (Bai et al. 2018). These policies have contributed to the situation that the amounts of N and P in manure, wastes and residues produced are larger than the demands of N and P by the crops grown in 30% and 50% of the number of counties, respectively. Thus, these policies have contributed to increased nutrient losses (Bai et al. 2018; 2021). In 2017, the central government implemented ‘none-livestock-production regions’, to protect eutrophication-sensitive water bodies and natural areas. This policy led to the shut-down of 0.26 million pig farms in ‘none-livestock-production regions’ and decreased the number of pigs in China by 46 million (Bai et al 2019). Further studies indicate that a drastic increase in the recycling and utilization of N and P from manure, wastes and residues can only happen following relocation of 5 to 10 billion of animals from south and east China to areas with sufficient cropland in the north and west. This would strongly reduce eutrophication of water bodies and the exposure of humans to high ammonia emissions (Bai et al. 2019; 2022).

Governmental agencies have a strong role in the agricultural supply and advisory system (Sect. " Box 1. Nutrient management"), but the agricultural agencies are more powerful than the environmental protection agencies, or were so before 2015 (Ma et al. 2013). The fragmented agricultural advisory system may also explain indirectly the success of the so-called Science and Technology Backyards, i.e., hubs that connect the scientific community directly with the farming community, to facilitate information exchange and stimulate innovations (Zhang et al. 2016; Cui et al. 2018). Since about 2009, nearly 1000 Science and Technology Backyards have been set-up, which potentially provide technological advice to farmers in about 600 counties. Rigorous statistical analyses of results of surveys on crop farms reveal indeed that the Science and Technology Backyards program increased grain yields, particularly in wheat production, and decreased fertilizer use, and thereby increased overall nutrient use efficiency (Li et al. 2023). The Science and Technology Backyard hubs have been able to reach-out to about 20 million farmers now (out of a total of about 250 million farmers), but it remains to be seen whether the farmers of the hubs achieve indeed environmentally-sound nutrient management practices and whether the potentials of best crop and animal husbandry practices are being achieved by all farmers.

In 2016, China’s government proclaimed the need for further restructuring the agricultural sector and the associated agricultural supply and advisory system through ‘Agriculture Green Development’. Agriculture Green Development is a response of China’s government and academia to the approval of the United Nations Sustainable Development Goals in 2015. It aims primarily at the transformation of current agriculture into a ‘green agriculture and countryside’ with high productivity, high resource use efficiency and with low environmental impact. It demands active collaboration between government, research, education, industry and advisors, and has four pillars, i.e., green crop production, integrated crop and animal production systems, green food processing, and enhancing the rural environment and ecosystem services (Shen et al. 2020). Thus far, it is largely a university-led and government-supported research program, but with the potential to generate structural reform and to contribute to improving nutrient management in practice (Zhang et al. in press).

Summarizing, China’s government has a large influence on agricultural development and nutrient management planning, mainly with the aim to achieve food security and sufficient farm income, but increasingly also to improve the rural environment. The government strongly supports the agricultural research of universities and the China Academy of Agricultural Sciences, who develop improved germplasm, breeds, techniques and new nutrient management concepts. It supports the regional agricultural bureaus, with hundreds of thousands of officers, who advise farmers and collect and analyze farm data. It supports crop farmers for growing certain crops, livestock farmers for investments in the modernization of animal houses and manure processing techniques, and the fertilizer industry for delivering enhanced and cheap fertilizers to farmers. More recently, governmental agencies force intensive crop and livestock farms to move away from eutrophication sensitive areas. The governmental interferences have a top-down character. The messages and incentives trickle down from the central government to the governments of the 32 provinces of mainland China, then to the officers of 687 cities and those of the 2850 counties, and finally to about 0.7 million villages, each with a village leader and 100 to 500 farm households.

A new perspective–farm-specific tools and techniques

Essentially, nutrient management is a cyclic process involving (i) analysis and assessment of nutrient demands and nutrient sources, (ii) decision making and planning of actions, (iii) careful implementation of actions, and (iv) monitoring, evaluation and refinement of planning and actions (Beegle et al. 2000). Nutrient management requires systems analysis, soil analyses, simple models for analyzing nutrient cycling and budgets, and best management practices that consider growth defining, limiting and reducing factors and emission mitigation principles (Van Ittersum and Rabbinge 1997; Sutton et al. 2022). These tools, techniques and farm-based nutrient cycling and budgeting models have to be developed by interdisciplinary teams, who understand farming in practice, nutrient cycling and management, technology transfer, and human behavior.

Currently, farmers in China have no tools and techniques for environmentally-sound precision nutrient management; they farm without the customary dashboard, indicators, steering wheel, brake, accelerator and clutch. Crop farmers mainly steer by choosing fertilizer bag size and nutrient formulation. Livestock farms often have no sufficient nearby cropland and incentives for proper manure utilization, and many crop farmers have no adequate information about manure quality and have no proper techniques for manure application (Zhang et al. 2022).

Thus, farm- and farmer-specific tools and techniques have to be developed and tested for different farm types. These tools have to be tailormade, so that the farmers understand and can deal with the steering wheel and the dashboard with indicators. These tools and techniques may also need regular updates, because some developments go really fast. For testing of the tools, there is a need for in-depth whole farming system studies, where nutrient cycling and management is monitored, measured, analyzed and synthetized over many years, also to understand the feedbacks of farm management on productivity, soil fertility and quality, nutrient use efficiency, and nutrient losses over time. Scientists have conducted many factorial field experiments, snap-shot farm surveys and, to a lesser extent, modelling studies during the last few decades, but long-term farm studies hardly exist in China. Such studies are needed because they allow to increase the understanding of the differences between farms in performance, through examination of the strategic, tactic and operational management of farmers over time, and would yield tested tools, techniques and (best) management practices.

Initial focus could be on entrepreneurial farmers in horticulture, fruit production, and livestock farming, because these farmers will benefit most from the tools, techniques and new insights generated. Also, the environmental impacts may be relatively large in these sectors. In crop production, the initial focus could be on enlarged farms, because larger farms tend to be more productive and use nutrients more efficient, and these farmers are likely more interested in adopting new tools and technology (Chen et al. 2022; Zhu et al. 2022). Following extensive testing and improving of the tools and techniques, other farmers should be supplied with appropriate tools and techniques as well.

A possible linkage of farm-based models with monitoring and accounting tools and incentives via the internet should be discussed further. It would also allow to reward good farm performance with premiums from food processing sectors and retail, or link governmental support to farm performance in a fast and transparent manner. However, there are privacy and confidentiality implications involved here, and the tools and models have to be sufficiently robust to allow credible reporting and accounting.

Importantly, the development of the tools and techniques should be linked to the concepts, ideas, and targets of the Agriculture Green Development program and policy, as this program and associated policy aim at a restructuring and greening of the agricultural and food processing sectors in China within the next decades (Shen et al. 2020). The scope of the Agriculture Green Development program is rather broad; it focuses on increasing productivity and economic profitability, land and water use efficiency, nutrient use efficiency, and soil carbon sequestration, and on minimizing greenhouse gas emissions, nutrient losses, pesticide use and energy use. A key pillar of Agriculture Green Development is enhancing ecosystem services, protecting biodiversity and restoring degraded landscapes (Shen et al. 2020; Zhang et al. in press). These different foci also require appropriate tools and incentives. Preferably, just one modeling/monitoring/accounting tool should be developed per farm type, to prevent overload. These tools have to be simple and accurate; but farmers may need the help of advisors for completing the input files and for interpretation of the results.

Our hypothesis is that farm-specific techniques and tools, and sustainability-driven farm business models are key to environmentally-sound nutrient management. Sustainability-driven farm business models are also a panacea to Agriculture Green Development. The implication of this hypothesis is that the businesses of farmers and other stakeholders should be considered as the starting point for the transition to environmentally-sound nutrient management and agriculture green development. The required farm business models are sustainability-driven, because farmers are rewarded for those functions that society demands, and that have impact on the environment and society. Thus, farmers have to be rewarded not only for the food, feed and fiber they deliver, but also for environmental-sound nutrient management, SOC sequestration, greenhouse gas mitigation, protection and provisioning of functional and intrinsic biodiversity, nutrient cycling, landscape maintenance, and water storage, purification and regulation. This requires a redirection of governmental support, and greater involvement of processing industry, retail and suppliers in the renumeration of farmers.

Setting-up such sustainability-driven farm business models will be a joined effort of farmers, industry, policy makers, scientists and other actors in the food production-consumption chain. Pilots could be established first at regional levels. Let us find out whether sustainability-driven business models are a panacea to environmental-sound nutrient management.

Conclusions

Nutrient management in China is at the crossroads. Decisions have to be made about the development and testing of farm-specific techniques and tools to provide farmers a customary dashboard with indicators and steering wheel for precision and environmentally-sound nutrient management. Decisions have to be made also about the development of sustainability-driven farm business models and integrated crop-livestock farming systems within the context of Agriculture Green Development, which aims at greening agriculture and enhancing ecosystem services, protecting biodiversity and restoring degraded landscapes. These tools, techniques and models have to be developed jointly by scientists, farmers, advisors, policy makers and food processing and retail.

Though fertilizer use has started to decrease from 2016 onwards, following a steep and continuous increase since the 1970s, total inputs of N and P in crop and animal production and the associated N and P losses are still large. Various factors may have contributed to the recent decline in fertilizer use, but there is no guarantee that these factors lead in the end to environmentally-sound nutrient management and sufficient production of nutritious food. As a matter of fact, there are several barriers in practice for environmentally-sound nutrient management. We argue that the most decisive barrier is that farmers currently have no appropriate information, tools, techniques and incentives for conducting precision and environmentally-sound nutrient management. Instead, most crop farmers manage primarily by choosing fertilizer bag size and nutrient formulation, while many livestock farmers have a headache of manure management, because of malfunctioning manure processing techniques and a lack of appropriate manure recycling opportunities.

Evidently, there is a need for increased crop-livestock integration (Hou et al. 2021). Regions with high livestock density will have to decrease livestock density to allow proper utilization of manure nutrients, and to minimize nutrient losses. Incentives for manure processing, transport and utilization have to be redirected to the end-users of manure (Tan et al. 2021). There is also need for simple manure nutrient accounting systems on both livestock and crop farms to be able to monitor progress.

Looking forward, researchers should be encouraged to conduct whole-farm analysis over a long term. They should develop and test new tools, techniques, green farming systems, sustainability-driven farm business models, together with entrepreneurial farmers, advisors from agricultural bureaus, policy makers, and food processing and retail. Instead of conducting another field experiment and/or another snap-shot farm survey, there is need for in-depth and long-term studies of whole farming systems to better understand the impact of farmers’ decisions on nutrient cycling, use efficiency and losses, to test and improve nutrient management tools, and to explore the effects of policy measures.

Such approaches could be applied and tested first in Science and Technology Backyards and then up-scaled to other villages and counties. It may involve also training and coaching of the advisors of the agricultural bureaus and enterprises, as these advisors have a key role in technology transfer and supporting farmers. These suggestions will pave the way towards accelerating the implementation of precision and environmentally-sound nutrient management in farms, and thereby contribute to increasing nutrient use efficiency and to decreasing nutrient losses in an accountable manner.