Soil health assessment is fundamental to guiding sustainable soil management practices, ensuring healthy soil, crop productivity, and provision of other ecosystem services. Interpretation of soil health in intensive agriculture in the North China Plain (NCP) is still lacking due to an over emphasis on soil chemical management and large variations among smallholders’ farmlands. The objectives of this study were to (i) compare soil health assessment approaches in response to fertilization regimes, (ii) quantify relationships between soil health and agronomic outcomes, (iii) develop a minimum data set to simplify soil health assessment, and (iv) validate soil health assessment frameworks in smallholders’ fields on the NCP. We collected soil samples from eight wheat-maize rotation long-term experiments which were divided into three fertilization regimes: (1) NPK, application of chemical fertilizers only; (2) M, application of organic materials only; and (3) MNPK, combined application of organic materials with chemical fertilizer application. Three soil health indexing (SHI) approaches, Comprehensive Assessment of Soil Health (SHI-CASH), linear (SHI-L) and sigmoidal (SHI-Sig) were evaluated. SHIs in the M and MNPK treatments were significantly higher than those in the NPK treatment across assessment approaches and were positively correlated with maize yield. A minimum data set including subsurface hardness, wet aggregate stability, available K, available Fe, soil organic carbon and soil protein was established using best subset regression. The soil health indices of smallholders’ farmlands using CASH and MDS were 0.58 (0.42–0.73) and 0.63 (0.40–0.87), respectively. More than 60% smallholders’ fields was at middle or low level. The relationship between SHI-CASH and SHI-MDS (Sig) was better than those for the CASH and linear methods. Our results demonstrate that an MDS based on best subset regression is applicable for evaluating soil health in wheat-maize rotation systems in the NCP. Soil health assessment in smallholders’ farmland indicates that soil health constraints are related to soil and biomass management, which provides insights on pathways towards addressing soil health gaps.

土壤健康评估是指导可持续土壤管理实践、确保健康土壤、作物生产力和提供其他生态系统服务的基础。由于过分强调土壤化学管理和小农耕地之间的巨大差异，华北平原 (NCP) 集约化农业土壤健康的解释仍然缺乏。本研究的目的是 (i) 比较土壤健康评估方法以响应施肥制度，(ii) 量化土壤健康与农艺结果之间的关系，(iii) 开发最小数据集以简化土壤健康评估，以及 (iv ) 验证 NCP 小农田间的土壤健康评估框架。我们收集了八个小麦-玉米轮作长期试验的土壤样品，这些试验分为三种施肥方案：（1）NPK，仅施用化肥；(2) M，仅应用有机材料；(3) MNPK，有机材料与化肥结合施用。评估了三种土壤健康指数 (SHI) 方法、土壤健康综合评估 (SHI-CASH)、线性 (SHI-L) 和 S 形 (SHI-Sig)。跨评估方法，M 和 MNPK 处理中的 SHI 显着高于 NPK 处理中的 SHI，并且与玉米产量呈正相关。最小数据集包括地下硬度、湿骨料稳定性、可用 K、可用 Fe、(3) MNPK，有机材料与化肥结合施用。评估了三种土壤健康指数 (SHI) 方法、土壤健康综合评估 (SHI-CASH)、线性 (SHI-L) 和 S 形 (SHI-Sig)。跨评估方法，M 和 MNPK 处理中的 SHI 显着高于 NPK 处理中的 SHI，并且与玉米产量呈正相关。最小数据集包括地下硬度、湿骨料稳定性、可用 K、可用 Fe、(3) MNPK，有机材料与化肥结合施用。评估了三种土壤健康指数 (SHI) 方法、土壤健康综合评估 (SHI-CASH)、线性 (SHI-L) 和 S 形 (SHI-Sig)。跨评估方法，M 和 MNPK 处理中的 SHI 显着高于 NPK 处理中的 SHI，并且与玉米产量呈正相关。最小数据集包括地下硬度、湿骨料稳定性、可用 K、可用 Fe、跨评估方法，M 和 MNPK 处理中的 SHI 显着高于 NPK 处理中的 SHI，并且与玉米产量呈正相关。最小数据集包括地下硬度、湿骨料稳定性、可用 K、可用 Fe、跨评估方法，M 和 MNPK 处理中的 SHI 显着高于 NPK 处理中的 SHI，并且与玉米产量呈正相关。最小数据集包括地下硬度、湿骨料稳定性、可用 K、可用 Fe、使用最佳子集回归建立土壤有机碳和土壤蛋白质。使用CASH和MDS的小农农田土壤健康指数分别为0.58（0.42-0.73）和0.63（0.40-0.87）。60%以上的小农田属于中低水平。SHI-CASH 和 SHI-MDS (Sig) 之间的关系优于 CASH 和线性方法。我们的结果表明，基于最佳子集回归的 MDS 适用于评估 NCP 小麦-玉米轮作系统中的土壤健康。小农农田的土壤健康评估表明，土壤健康制约因素与土壤和生物量管理有关，这为解决土壤健康差距的途径提供了见解。