Fallow priority areas for spatial trade-offs between cost and efficiency in China

Agriculture must ensure sufficient, sustainable food production while curbing environmental degradation, including soil pollution and groundwater overdraft, linked to intensive practices. Fallowing—temporarily halting cultivation—has long been used to reduce negative externalities and restore soil function. In China, fallow is a key policy instrument to protect cultivated land, reduce soil pollution, and support sustainable agricultural development and food security. However, national guidance to implement fallow with explicit cost–benefit trade-offs is lacking, and it remains unknown where fallow can deliver the greatest ecological and food safety benefits at the least cost. The study addresses this gap by identifying where fallow implementation can be most cost-efficient given four eco-environmental stressors (soil pollution, groundwater overexploitation, cultivated land quality, and ecological protection redlines) and by evaluating spatial trade-offs among pollution control benefits, environmental protection benefits, and implementation costs under a realistic upper fallow limit of 20% of cultivated land.

Prior work on spatial trade-offs has optimized land-use decisions for ecosystem services, biodiversity conservation, restoration, and protected areas, but national-scale prioritization of fallow areas with explicit cost–benefit analyses has been missing. Existing studies in China primarily estimated fallow scales from food security or population-carrying perspectives without spatial prioritization, or analyzed regional fallow based on single aspects due to data limitations. For example, studies projected fallow ratios under future food security constraints and calculated theoretical fallow scales from land-use change and consumption trends, yet they did not map where fallow should occur, nor did they evaluate multi-objective trade-offs or costs. China’s current bottom-up, voluntary fallow declarations and total scale control lack guidance on urgency across eco-environmental stressors and on cost-effective spatial targeting. This study builds on these gaps by integrating multiple stressors and costs to produce national-scale, scenario-based spatial prioritizations.

Data and indicators: The study compiled multisource national datasets: (1) soil pollution mapped from 553 peer-reviewed studies covering 1,781 sites and 5,597 samples, assessing eight key pollutants (Cr, Cu, Cd, Pb, Zn, HCHs, DDTs, PAHs); (2) cultivated land quality grades (2015 MLR) aggregated to five classes (excellent to inferior), extended to 2020 arable land using 30 m land-cover data and neighborhood assignment; (3) groundwater overexploitation zones derived via the water level amplitude method comparing 1979 vs 2011 shallow groundwater levels and classified into balanced, slight, moderate, and severe, validated with national atlases and guidelines; (4) ecological protection redlines (EPR) delineated by combining ecosystem services importance (water retention, soil conservation, biodiversity maintenance) and environmental sensitivity (desertification, soil erosion, rocky desertification), cross-checked with 20 provincial EPR releases and classified into first-class, second-class, and outside EPR. Soil pollution assessment: Single factor and Nemerow integrated pollution indices were used; carcinogenic risk (CR) to children was calculated via ingestion, inhalation, and dermal pathways using USEPA frameworks, with BaP representing PAH carcinogenicity. Acceptable CR threshold was set to 1×10^-5. Cultivated land quality, groundwater overexploitation, and EPR were reclassified to ordinal scales (Q: 0–8; G: 0–9; E: 0–12) reflecting increasing urgency/benefit if fallowed. Objective functions and costs: Three objectives were formulated: (1) maximize pollution control benefit RI based on standardized CR reduction; (2) maximize environmental protection benefit EP based on the sum of standardized Q, G, E improvements upon fallow (restoring quality to at least Grade 8, groundwater to balanced, and EPR lands to the condition of outside-EPR); (3) minimize fallow implementation cost CT, comprising: lost production (P = y × x, where spatial yield potential y and average crop income x = 0.177 USD kg^-1), farmer subsidy S equal to regionalized land transfer price, fallow matching funds F = 382.5 USD hm^-2 (technical training, data services, green manure seeds and grants, mechanical shredding/return, microbiological inputs), and management costs M differentiated by stressor severity (e.g., phytoremediation costs for heavy metals). Multi-criteria optimization: A spatial optimization identified priority fallow areas by maximizing or minimizing combinations of standardized RI, EP, and CT across cultivated land grid cells, with decision variable x_i indicating the fallow proportion per grid. The optimization considered scenario-specific objective configurations and weights (w_RI = w_EP = 0.5 in multi-benefit and comprehensive scenarios). Fallow scale constraint: To align with food security, total fallow area was constrained to a maximum of 20% of national cultivated land, consistent with recommended self-sufficiency thresholds under future demographic and consumption trajectories. Scenarios: Five scenarios were analyzed—(1) PMS: maximize RI only; (2) PES: maximize EP only; (3) CRS: minimize CT only; (4) MBS: maximize RI and EP (w_RI = w_EP = 0.5), no cost; (5) CFS: maximize RI and EP (w_RI = w_EP = 0.5) while minimizing CT. Spatial analyses and raster calculations were conducted in ArcGIS 10.6; sensitivity analyses addressed pollution data outliers using the [x/4, 4x] approach; mapping resolution was 1 km².

- Baseline stressors: 77.9% of cultivated land is unpolluted, 18.6% slightly polluted, and 3.5% moderately to severely polluted, concentrated largely south/east of the Hu Huanyong line. Only 29.5% of arable land is good or excellent; 17.7% is poor or inferior. Groundwater is balanced on 91.8% of cultivated land; 4.5% faces moderate to severe overexploitation, notably in Hebei and Henan (~4670 and ~3950 km²). About 14.5% of cultivated land lies within EPR (3.6% first-class, 10.9% second-class). - Spatial priority patterns vary markedly by objective: PMS priorities concentrate in southeastern China (moderate/severe pollution hotspots); PES priorities in northern/central regions; CRS priorities in northern/western China. MBS concentrates around the Loess Plateau hilly/gully areas (Gansu, Henan–Hebei junction, SE Inner Mongolia, S Anhui). CFS resembles CRS in dispersion but with higher priority density in the northeast at larger fallow shares. - Outcome variations at small fallow shares (2.1%): For the same fallow percentage, pollution control benefit varies by 95.0 times (α = 23.5%, from 0.3% in CRS to 23.8% in PMS). Environmental protection benefit varies over eightfold (β = 9.9%, from 1.3% CRS to 11.2% PES). Total cost for 2.1% fallow differs by ~192.8 billion USD among scenarios; cost ranges (billion USD): PMS 152.2–242.7; PES 29.4–41.3; CRS 3.9–8.7; MBS 138.2–256.9; CFS 8.5–20.7. - At 20% fallow (upper limit): Maximum pollution control benefit reaches 98.7% (PMS); maximum environmental protection benefit reaches 64.7% (PES), exceeding PMS, CRS, MBS, CFS by 42.4%, 48.1%, 32.5%, and 38.1% respectively. Costs vary 6.7× at minimum and 7.4× at maximum across scenarios, with CRS costing 56.6–79.4 billion USD and PMS 380.4–588.7 billion USD. - Comprehensive trade-off (CFS) at 20% fallow: Achieves up to 28.8% pollution control benefit and up to 26.6% environmental protection benefit at a total cost of 65.5–95.6 billion USD, implying savings versus PMS of 323.8–509.3 billion USD (up to 13.6% of China’s 2021 public budget). - Spatial congruence: PMS priorities align with heavy metal contamination zones (e.g., Hunan), and PES priorities align with groundwater overdraft areas (e.g., Henan), consistent with existing pilots.

The study demonstrates that where fallow is implemented crucially determines both environmental outcomes and fiscal costs. By explicitly modeling trade-offs, the authors show that identical fallow proportions can yield wildly different benefits (up to 95× for pollution control and over eightfold for environmental protection at low fallow shares), underscoring that area alone is an ineffective planning metric. The comprehensive scenario (CFS) balances pollution control, environmental protection, and cost, enabling substantial benefits at comparatively low national expenditure, thereby freeing budget for other sustainable agriculture targets. The results address the central question—where to fallow for maximal cost-efficiency—by providing spatially explicit priorities under different policy objectives and constraints, supporting China’s soil pollution mitigation, groundwater recovery, and ecological protection goals. The findings advocate for multi-objective, location-specific fallow planning to guide national programs, reduce unsustainable land use, and enhance implementation efficiency under limited funds.

This work provides the first national-scale, multi-objective spatial prioritization of fallow in China that jointly considers pollution control, environmental protection, and implementation cost under a realistic 20% fallow cap. The approach identifies priority areas that can deliver up to 98.7% pollution control benefit or 64.7% environmental protection benefit, and, under a comprehensive trade-off, achieve meaningful dual benefits at markedly reduced costs (saving up to 509.3 billion USD). The framework and data can guide national-to-local fallow planning, monitoring, and optimization of spatial-temporal allocations. Future research should integrate higher-resolution and continuously updated pollution data, incorporate social dimensions such as farmer willingness and behavioral responses, and operationalize fallow modes and timeframes to tailor interventions to local contexts while tracking long-term ecological and agronomic outcomes.

Two main caveats are noted: (1) Soil pollution datasets may include outliers and resolution constraints that could influence benefit estimates and priority mapping; sensitivity checks ([x/4, 4x]) were performed, and a 1 km² mapping resolution—smaller than typical village administrative units—was used. (2) Social factors, including farmer willingness to participate and associated behavioral responses, were not modeled due to lack of suitable data, potentially affecting practical uptake. Additionally, the study identifies where to fallow but does not specify how (fallow modes, duration), and cost estimates rely on available pilot-derived parameters that may vary regionally.