Soil and Its Interaction with the Climate Jointly Drive the Change in Basic Soil Productivity under Long-Term Fertilizer Management

Basic soil productivity (BSP) is the productive capacity of farmland soils with their own physical and chemical properties during a specific crop season under local field management. Improving BSP as an effective way to increase or maintain crop yield, and researching its changes and potential driving factors under long-term fertilization are crucial for ensuring high and stable crop yields. In this study, the yields of BSP were simulated using the decision support system for agrotechnology transfer (DSSAT) crop model based on three long-term experiments, and its changing characteristics and driving factors were investigated under various fertilization treatments in wheat–maize rotation systems during 1991–2019. Five treatments were included: (1) unfertilized control (CK); (2) balanced mineral fertilization (NPK); (3) NPK plus manure (NPKM); (4) high dose of NPK plus manure (1.5NPKM); and (5) NPK plus crop straw (NPKS). This study found that the BSP of wheat and maize exhibited a fluctuating increase or stable change trend under four fertilization treatments at the Yangling (YL) and Zhengzhou (ZZ) sites, while a fluctuating reduction trend was observed at the Qiyang (QY) site. Compared with CK, NPKM, 1.5NPKM and NPKS significantly (<i>p</i> < 0.05) improved the BSP of wheat and maize at the YL and ZZ sites. The BSP of both maize and wheat under NPKM and 1.5NPKM was significantly (<i>p</i> < 0.05) higher than that under NPK or NPKS, whereas no statistically significant difference was found between NPK and NPKS at the QY site. The contribution rates of basic soil productivity (CBSP) of maize and wheat at the YL (41.5–60.7% and 53.0–64.3%) and ZZ sites (44.4–59.2% and 56.8–66.7%) were overall higher than that at the QY site (25.8–37.6% and 48.3–60.1%). In most cases, the difference in CBSP among different fertilization treatments was consistent with that in BSP. Moreover, a significant positive correlation was recorded between BSP and soil pH (r = 0.73, <i>p</i> < 0.01; r = 0.87, <i>p</i> < 0.01), TN (r = 0.56, <i>p</i> < 0.01; r = 0.62, <i>p</i> < 0.01) and TK content (r = 0.49, <i>p</i> < 0.01; r = 0.58, <i>p</i> < 0.01) in maize and wheat. Soil pH significantly correlated positively with the BSP of maize (R² = 0.54, <i>p</i> < 0.001) and wheat (R² = 0.49, <i>p</i> < 0.001) at the QY site, but negative correlation (R² = 0.20, <i>p</i> < 0.001; R² = 0.30, <i>p</i> < 0.001) was only found in maize at the YL and ZZ sites. The BSP of maize and wheat showed a significant negative linear correlation with MAP (R² = 0.49–0.67, <i>p</i> < 0.001) and MAT (R² = 0.36–0.62, <i>p</i> < 0.001). Random forests (RF) and variance partitioning analysis (VPA) revealed that soil properties and its interaction with the climate showed a higher explanation rate for BSP, indicating that these factors are the key drivers of BSP change. Overall, chemical fertilizers combined with manure can effectively increase BSP, while the effects of fertilizer combined with straw on BSP vary by region. The changes in BSP in wheat–maize cropping were mainly driven by both the soil and its interactions with the climate.