Identifying where Japanese agriculture is most at risk: A longitudinal analytical framework based on municipal boundaries as of 1950 for workforce decline and aging (2005–2020)

Japan's agricultural workforce is shrinking and aging, posing a significant social issue. Yet, fundamental analysis-quantifying and mapping where and how rapidly this demographic shift is progressing-have been lacking, largely due to extensive municipal boundary reorganizations in Japan. This study aimed to visualize and clarify the current demographic shifts by restoring temporal comparability through a stable spatial baseline: the "sub-municipalities." Using Census of Agriculture and Forestry data for 2005, 2010, 2015, and 2020, we mapped two indicators for core agricultural workers: (i) the decline rate between 2005 and 2020, and (ii) the proportion of workers aged 75 years and above in 2020, revealing the regional landscape of workforce shrinkage and aging. We produced nationwide maps at the sub-municipality level, summarized land-type trends using generalized mixed models, and identified areas of extreme change using hotspot/coldspot analysis. The results revealed a nationwide downturn with pronounced spatial heterogeneity: the strongest declines were observed in mountainous and upland-dominated areas, whereas they tended to be more moderate in flatland paddy areas. Hotspots were scattered throughout the country, but were mainly located in areas with significant geographical constraints. Coldspots, representing modest growth, were identified around the Kinki and northern Kita-Kyushu metropolitan fringes. The aging rate was the highest in mountainous and paddy areas, whereas flatland and upland-dominated areas tended to be more resilient in this aspect. Aging hotspots aligned with the Tokai-Tosan Mountain belt and the Sanyo and San'in regions, whereas coldspots were observed in Hokkaido and Tohoku regions. Although we focused on the numerical and age composition of core agricultural workers, the approach can be generalized to other census indicators (e.g., sales, cultivated area, and production type), supporting locally adapted, evidence-based rural policy.