Nationwide big data analysis of the influence of the distribution trend of lowland soils on paddy yields in Japan

This study investigated the relationship between soil type and paddy yield in Japan on a national scale. Previous studies have primarily relied on field experimental data and were limited to site comparisons within regions under similar climatic conditions, making it challenging to clarify the influence of soil factors on paddy yields independently of climatic factors. To address this issue, we conducted a big data analysis using nationwide databases to statistically examine the relationship between soil type, climate, and paddy yield. A generalized additive model was used to evaluate the effects of soil distribution patterns and climatic conditions on paddy yields, based on municipal-level databases from 1993 to 2019. Soil distribution patterns within each municipality were categorized into five classes according to the proportion of the distribution area of lowland soil groups. In addition to soil distribution and climatic conditions, the model included prefecture and year as explanatory variables to account for fixed effects related to rice variety, management practices, and annual variation. The results showed that paddy yields tended to be higher in municipalities with a large distribution area of the Gray lowland soil group compared to those with other lowland soil groups. This finding reinforces the existing understanding of the influence of soil drainage conditions on crop growth and highlights the importance of soil management and assessment in rice cultivation. Furthermore, aggregating the distribution trends of soil types at the municipal level proved to be an effective approach for evaluating crop productivity at the national scale from the perspective of soil conditions. This study underscores the value of soil map information by clarifying its relationship with productivity and proposing strategies for its effective utilization.