In 2015, China consumed more chemical fertilizers than the United States and Europe combined. Between 2007 and 2017, the Nature Food study documents a 10% increase in yields associated with an 8% reduction in nitrogen pollution. This result is the product of a state program piloted province by province, with an official target by 2030 of achieving more than 43% fertilizer use efficiency for the three main cereals. The stakes extend beyond China: the data document a simultaneous improvement in yields, nitrogen efficiency, and nitrogen pollution, not a complete decoupling from fossil-based inputs.
The essentials
- Official figures indicate that in 2019 agricultural fertilizer use was 6.19 million tons lower than in 2015, with yields maintained (Nature Food, 2026).
- The state program Fertilizer Zero Growth, piloted by the Ministry of Agriculture, substituted biological inputs for synthetic fertilizers through territorial coverage by province.
- The 2025-2030 program sets targets for use efficiency and scientific fertilization, drawing on biological nitrogen fixation and precision technologies.
- China has substantial reserves and is the leading producer of phosphate rock; the program does not explicitly address phosphorus supply security, a question that cannot, however, be equated with simple import dependence.
- Biological intensification constitutes a credible pathway for large-scale agriculture, subject to verification on crops and zones poorly suited to biological nitrogen fixation.
The program that changed scale
In 2015, Beijing launched the Fertilizer Zero Growth program with a simple objective formulated against the entire logic of 20th-century agriculture: produce as much, with less chemistry. The Ministry of Agriculture asked agricultural administrative levels to break down targets annually and locally, trained technical advisors, and supported fertilization services and organic practices. Advice and fertilization based on soil analysis constituted one important lever among several tools in the program: the ministry simultaneously supported fertilization services, organic practices, and trained technical advisors.
Chinese authorities reported a 13.8% reduction in fertilizer use between 2015 and 2021. The available Nature Food study documents for 2007-2017 an increase in yields associated with management practices, but does not report this national reduction of 21%.
The method deserves as much attention as the figure. The program simultaneously coordinated public agricultural research, provincial institutes, agricultural cooperatives, and individual farmers. Each province had its own targets, adapted to its dominant crops and soil quality. This architecture, decentralized in execution and centralized in target, distinguishes the program from a simple national directive that remained a dead letter.
The reasons for maintained yields
A mechanism often underestimated in this type of transition is cumulative learning effects among trained farmers. When a farmer learns to read visual indicators of nutritional deficiency or excess, he adjusts his practices from one season to the next without waiting for external advice. This gradual adjustment reduces losses from runoff and volatilization, two pathways through which a significant share of applied fertilizers were lost without agronomic benefit. Training thus becomes a sustainable input, whose effects accumulate over time rather than being depleted like a chemical application.
The simplistic thesis would be that Chinese farmers were over-fertilizing, and that it was enough to remove the excess to obtain the same result. The agronomic reality is more nuanced. Nature Food attributes the observed changes to several management practices, including organic and chemical fertilizers combined, straw recycling, and deep incorporation.
The documented practices include the combination of organic and chemical fertilizers, which optimizes inputs; recycling of straw and other organic waste, which recycles nutrients; and deep incorporation as well as improved application methods, which reduce losses from runoff. CGIAR is studying nitrogen-fixing bacterial strains best adapted to Chinese agro-climatic conditions.
The improvement in yields and nitrogen efficiency is associated with this set of practices. This is systematic, financed input substitution. The robustness of this result remains to be verified on a larger scale and on crops less suited to biological nitrogen fixation, such as wheat or corn in the semi-arid zones of the north of the country.
The logic of territorial investment
Economist Olivier Lluansi, in his work on French reindustrialization, defends a thesis that illuminates the Chinese model in an unexpected way. According to him, major industrial transitions fail when they remain abstract national policies, without strong territorial anchoring: without mobilizing local actors, without directing funding toward the basins concerned, without regional institutions capable of adapting the strategy to ground realities.
The Fertilizer Zero Growth program illustrates exactly this principle applied to agriculture. The central state set and coordinated the objective; the scale and source of funding are not precisely established. Provinces adapted targets to their crops and supervisory capacity. Local agronomic institutes trained technicians. Cooperatives relayed information to farmers.
Without this chain of territorial actors, the program would have been a circular with no effect.
The question of financing that Lluansi raises for French reindustrialization—namely, who finances the decade of transition between the central state, regions, and private savings—arises in an analogous way for China. The program benefited from substantial public support, including central funding for the fertilization program based on soil analysis. The official plan sets a 2030 target without an additional 20% reduction target, and relies on the integration of new techniques and improved fertilizer efficiency. The distribution of financing becomes less obvious once the most accessible gains have been achieved.
A strategic vulnerability that the transition does not erase
This structural dependence on imported phosphorus reveals an asymmetry in the program’s very design: the biological levers mobilized target nitrogen because solutions exist, not because nitrogen represents the most critical constraint over the long term. Piloting a transition based on available solutions rather than actual vulnerabilities is a politically understandable logic, but it introduces a bias into the assessment of results. What the program measures as success does not encompass all the fragilities that the transition was supposed to reduce.
The reduction of nitrogen and potassium fertilizers is one thing. Dependence on phosphorus is another, and the Chinese program does not solve it. China is a major producer of phosphate and its supply situation depends on its production and substantial national reserves. Phosphorus is a resource with no direct biological substitute: no bacterium fixes atmospheric phosphorus, because there is no vast gaseous phosphorus reservoir comparable to that of dinitrogen. It must be extracted from soil or recycled.
This vulnerability is geopolitically significant. Morocco alone holds approximately 68% of estimated global phosphate rock reserves. China, which aspires to strengthened food self-sufficiency, possesses a major domestic base of phosphate production and reserves. The 2025-2030 program does not address this problem head-on. The program improves overall fertilization efficiency, with explicit attention to nitrogen management, but leaves the phosphorus question entirely open.
This point limits the scope of the model for other countries wishing to replicate it. A complete agronomic transition requires addressing all three essential macronutrients. On two of three, China is moving forward. On the third, it remains exposed, and the program’s silence on this subject deserves to be noted.
This relates to a broader challenge facing all economies engaged in ambitious industrial transitions: structural dependencies on imported resources are not resolved by a single program, however well designed. Agricultural value chains have their own chokepoints, and phosphorus is one of them.
The 2030 objective and its conditions for success
The 2025-2030 program presents larger scaling challenges than the first phase, notably on less flexible agricultural systems. The first gains are always the most accessible: they correspond to the most manifest over-fertilization, to soils richest in residual nutrients, to farmers most receptive to training. The second tranche of reduction affects more constrained systems, less flexible crops, farmers harder to convince.
The official program relies on the integration of new techniques, products, and equipment, as well as on soil analysis-based fertilization and the promotion of organic fertilizers and formulated fertilization.
These approaches are being tested in pilot and experimental contexts. The challenge is scaling up to farmers whose average size remains small and whose digital connectivity is uneven across provinces. CGIAR is following this expansion with standardized measurement protocols: its data on actual adoption rates by farmers will be the most reliable indicators for assessing whether the 2030 target is achievable.
The limits and potential of programmed biology
The program raises a broader question about the possibility of massively substituting biological inputs for synthetic fertilizers without compromising yields, an issue that extends beyond China and interests other large-scale agriculture sectors.
The answer depends on two variables that evolve at different rates. The first is science: biological nitrogen fixation is advancing rapidly. Research teams in the United Kingdom and the United States are testing genetically modified cereal varieties designed to harbor nitrogen-fixing bacteria in their roots, a natural property of legumes, transposed experimentally to wheat and rice. If these varieties passed regulatory steps and reached farms by 2035, the biological ceiling would rise significantly.
The second variable is institutional: the capacity of agricultural systems to absorb these new technologies on a large scale. A farmer cultivating two hectares in Sichuan province does not have the same resources as a Midwest American farm to test, adopt, and finance innovations. The actual adoption rate of biostimulants in the second phase of the Chinese program will be an indirect measure of this absorption capacity. The actual adoption rate in the least-equipped provinces will be a key indicator of the system’s capacity to generalize gains.
The Chinese model combines public coordination, input efficiency, and production maintenance; it tests a territorial and integrated approach to agronomic transition. It is an agronomic and institutional wager, not yet a certainty. CGIAR data on the second phase of the program will be decisive for assessing the reproducibility of results on a larger scale.
At the global level, the same financing problem arises as in industrial transitions in other sectors: who finances the decade of transformation before gains become visible. In China, the answer has been the central state. In other contexts—India, Brazil, sub-Saharan Africa—neither budgets nor territorial institutions permit this type of integrated program. The model is instructive; it is not directly transposable.
Sources
- Nature Food (2025), Fertilizer reduction and yield stability in China’s agricultural transition
- Ministry of Agriculture of the People’s Republic of China, National Fertilizer Zero Growth Program 2015-2022 and objectives 2025-2030
- CGIAR, Data on adoption of biological nitrogen fixation practices and yield monitoring by province
- Olivier Lluansi, “Olivier Lluansi reveals the main lines of his targeted path for reindustrialization,” Banque des Territoires
