Spatiotemporal dynamics of maize yield and associated driving factors in the black soil region of Northeast China

The black soil region of Northeast China can serve as a crucial grain production base in the world. However, there is a complex interaction between the climate and soil fertility. This present study aims to explore the spatiotemporal dynamics of the maize yield and its key driving factors in the black soil region. The monitoring climate data was collected from the Cultivated Land Quality Monitoring Network of the Ministry of Agriculture and Rural Affairs from 1988 to 2021. Shapley Additive exPlanations (SHAP) interpretable machine learning model was employed to quantify the influence of the soil and climate factors on the maize yield. Partial least squares regression was used to determine the coupling relationship between the maize yield and climate or soil factors. The contribution rate of the soil fertility was finally calculated for the high yield. The results showed that the annual maize yield increased from 8.09 t/hm² during Stage I (1988–2000) to 8.83 t/hm² in Stage II (2001—2010), and further to 10.15 t/hm² in Stage III (2011—2021). The contribution rate of the soil fertility rose from 47% in Stage I to 53% in Stage II and 59% in Stage III. Spatially, the maize yield and soil fertility contribution rates were higher in the eastern region (Songnen Plain on the eastern of the Greater Khingan Mountains, with averages of 11.75 t/hm² and 82%, respectively), and lower in the northwestern region (northwestern Inner Mongolia, with averages of 6.95 t/hm² and 34%, respectively). Climate factors, particularly the drought index and diurnal temperature range, were identified as the primary drivers of the maize yield, accounting for 61.34% of the relative contribution. Soil fertility indicators, notably available phosphorus and available potassium, were also key controlling factors on the yield, thus relatively contributing 38.66% of the total. The SHAP analysis revealed that the available phosphorus and potassium promoted the maize yield. Their positive correlation was saturated relatively, when the available phosphorus and potassium exceeded 18 and 170 mg/kg, respectively. The drought index of 0.5 to 1.5 and the diurnal temperature range of 9 to 11℃ were found to be optimal for the maize yield. Interaction analysis showed that there was a positive correlation between the higher available phosphorus (＞18 mg/kg) with the lower drought index (0.5-1.5) and suitable diurnal temperature range (9-11℃). Partial least squares regression indicated that the climate and soil factors together were used to explain 24% of the spatiotemporal variation in the maize yield. Climate factors shared a strong direct effect on the maize yield (path coefficient of 0.43). While their indirect effects were weaker through soil nutrients and pH. Among the climate factors, the relative humidity, heavy rainfall days, and extreme rainfall shared the positive effects, whereas the negative effects were found in the warm spell duration index, drought index, and diurnal temperature range. In total, the climate factors shared a stronger influence on the maize yield (45.74%), compared with the soil nutrients (16.48%) and pH (4.69%). This finding can provide the theoretical support to predict the dynamics of the maize productivity in the black soil region of Northeast China under future climate and soil fertility. It highlights the climate factors monitoring and soil improvement measures for the stable and high maize yields. Future research should explore the interactions between climate, soil factors, and agricultural management practices. A more comprehensive understanding of the maize yield dynamics can be gained to support sustainable agriculture.