Topics

Exploring factors contributing yield variability of Tamba black soybean using historical cultivation data

- Toward a new water management guideline for stable production -

Updated:February 10, 2026 (Tuesday)

The National Agriculture and Food Research Organization (NARO) analyzed historical cultivation data accumulated over 16 years from farmers' fields in Tambasasayama city to investigate the factors behind yield fluctuations of Tamba black soybeans. This analysis employed both the Soybean Irrigation Support System developed by NARO and a newly developed, machine learning-based yield prediction model. The results suggested that soil moisture management in early October plays an important role in improving and stabilizing yield. This finding is expected to serve as a basis for proposing new water management guidelines. The study also suggests that the possibility of combining historical data from other regions with the Soybean Irrigation Support System to optimize region-specific crop management practices.

Overview

Tambasasayama city has a history of cultivating black soybeans for over 300 years, and its traditional farming system has been recognized as a Japanese Nationally Important Agricultural Heritage System. However, large fluctuations in yield have been a challenge for farmers. Since high temperatures and dry conditions are often encountered during the flowering stage, timely and sufficient irrigation is considered important to improve yields. In Tambasasayama city, the combination of an inland climate, clay-rich impermeable soils, and basin-shaped terrain makes farmland prone to both waterlogging and drought. Therefore, careful management of soil moisture is essential to stabilize black soybean yields.

NARO analyzed the factors contributing to yield variability of Tamba black soybeans using 16 years of cultivation data collected from farmers' fields in Tambasasayama city. Soil moisture over this period was estimated using the Soybean Irrigation Support System that incorporates data from the Agro-Meteorological Grid Square Data developed by NARO. This web-based tool, which was made publicly available in 2022, estimates soil moisture in real time and notifies users of optimal irrigation timing. By combining soil moisture estimates and weather data with a newly developed, machine-learning-based yield prediction model, the analysis revealed that temperature, solar radiation, and soil moisture during the flowering to seed-filling stages (August to October) have a complex influence on yield. In particular, the results suggest that appropriate soil moisture management in early October may play an important role in stabilizing yields.

While the Soybean Irrigation Support System has primarily been used to guide water management for conventional soybean cultivars, this study is the first to demonstrate its applicability to black soybeans. The results suggest that the system can be extended to a broader range of soybean cultivars. By applying similar approaches in other regions and accumulating historical cultivation data, it may be possible to develop region-specific water management guidelines to help stabilize yields. These efforts represent an important step toward building sustainable agricultural systems that are resilient to climate change.

Reference Information

Kumagai E, Minato M and Takahashi T (2024) Multifaceted analysis of the impact of recent changes in climate and soil moisture on the seed yield of Tamba black soybeans in Tambasasayama City. Japanese Journal of Crop Science, 93(4), 278-293.
https://www.jstage.jst.go.jp/article/jcs/93/4/93_278/_pdf
The main text is in Japanese, with the abstract available in English.

For Inquiries

Contact: https://www.naro.go.jp/english/inquiry/index.html

Centers・Institutes