Fertilization alters crop yield, soil aggregation, and carbon sequestration potential. However, the specific effects and interactive mechanisms of long-term fertilization on soil organic carbon (SOC), aggregate-associated organic carbon (OC), and yield in dryland wheat fields remain poorly understood. Therefore, a field experiment with local farmer fertilization (NP), measured and controlled fertilization (NPK), chemical fertilizer plus organic fertilizer (NPKM), chemical fertilizer plus bio-organic fertilizer (NPKB), and no fertilizer (CK) was conducted for 10 years in a semi-arid region of China. We examined the characteristics of crop yield, SOC stock, and soil aggregate under different fertilization methods to explore the carbon sequestration mechanisms associated with increased yield. The results revealed a significant positive linear correlation between carbon sequestration and carbon input, with a carbon sequestration efficiency of 27.4%. Different fertilization treatments increased the proportion of macroaggregates (>0.25 mm) and significantly improved the water stability of soil aggregates, with the combined application of organic–inorganic fertilizers (NPKM and NPKB treatments) showing the most pronounced effect. A redundancy analysis revealed that carbon input was the most important factor affecting aggregate stability. Partial least squares path modeling revealed that SOC content was enhanced primarily because of the increase in macroaggregate-associated OC and carbon input, which in turn led to higher wheat yields. Our findings indicate that macroaggregate-associated OC and carbon inputs have significant implications for SOC sequestration in dryland fields. Thus, we suggest the integrated application of organic fertilizers with balanced NPK fertilizers to promote SOC accumulation, improve aggregate stability, and enhance crop yield.

The Loess Plateau in China has long grappled with issues such as chronic soil erosion, poor soil structure, and diminished organic matter. Soil organic carbon plays a crucial role in enhancing soil fertility, and fertilization is a key tool that influences it. In a comprehensive field experiment, we examined five fertilization methods: no fertilization (CK); traditional fertilization used by local farmers (FP); nutrient-balanced fertilization with nitrogen, phosphorus, and potassium (OF); chemical fertilizers used in conjunction with organic fertilizers (OFM); and chemical fertilizers paired with bio-organic fertilizers (OFB). Our findings revealed that the OFM and OFB treatments were the most effective, explaining 84.35% and 81.26% of the variation in soil carbon sequestration, respectively. Further, the OF, OFM, and OFB treatments demonstrated superior effectiveness compared with the FP treatment in enhancing the soil carbon fractions. However, these fertilization patterns did not significantly alter the active-to-inert carbon ratio of the soil. OF, OFM, and OFB treatments enhanced the stability of soil carbon pools more than FP treatment. In structural equation modeling, factors such as microbial biomass nitrogen and phosphorus, soil pH, and β-N-acetyl glucosidase indirectly exhibited a limiting effect on the carbon pool stability index (CPSI), while β-glucosidase displayed an indirect positive correlation with the carbon fractions. In contrast, dissolved organic carbon, low-molecular-weight organic carbon, high-molecular-weight organic carbon, and crop yield demonstrated direct positive correlations with the CPSI. Consequently, both a balanced application of chemical fertilizers and the combination of organic and chemical fertilizers were effective in enhancing and sustaining the soil’s organic carbon content, thereby contributing to soil fertility stability. To this end, this study can inform the efficient selection and use of fertilizers, improving crop yield and soil carbon content.