Global Cs fallout inventories of forest soil across Japan and their consequences half a century later.

Japanese forests were exposed to multiple sources of radioactive contamination. To acquire scientific guidance on forest management planning, it is crucial to understand the long-term radiocesium (Cs) distribution (and redistribution) over time. To obtain robust evidence of the residual global fallout of Cs (Cs-GFO) after a few decades, we determined Cs-GFO inventory in forest soil at 1171 soil pits of 316 plots evenly spaced across Japan from 2006 to 2011, shortly before the Fukushima Dai-ichi Nuclear Power Plant accident. The activity concentration measurements were performed using a NaI well-type scintillation counter. The average (±SD) Cs-GFO in forest soil (0-30 cm from the surface) of the National Forest Soil Carbon Inventory (NFSCI) sampling plots uniformly extracted from the entire country was estimated to be 2.27 ± 1.73 kBq m (n = 316) as of Oct. 1, 2008. A high nationwide spatial variation was found in Cs-GFO, where relatively high Cs-GFO was found along the Sea of Japan compared with the total annual precipitation. We also obtained a reconstructed decay-corrected cumulative Cs-GFO dataset from the fallout observatories as the initial Cs-GFO. The cumulative Cs-GFO of fallout observatories averaged 2.47 ± 0.95 kBq m (n = 39) as of Oct. 1, 2008 and displayed spatial variation similar to that in forest soil. To identify whether Cs-GFO remains in forest soil across Japan, we examined a general linear mixed-effect model comparing Cs-GFO between forest soil and the observatory under normalized annual precipitation and region. The model did not indicate a significant difference, but relatively lesser Cs-GFO was found in forest soil, where the least-squares mean of Cs-GFO in forest soils was 79.1% of that of the observatory. The variation in Cs-GFO in forest soils within NFSCI sampling plots was 1.4 times greater than that among plots. The high spatial variation in Cs-GFO within a 0.1-ha plot strongly suggested the redistribution of Cs-GFO within the forest catchment. The vertical distribution pattern of Cs-GFO across three depth layers indicated that the Cs-GFO redistributions were likely attributed to the movements of sediments and mass. Moreover, when extracting soil pits assumed to have the least soil disturbance from the vertical distribution pattern, no significant difference in Cs-GFO was observed between forest soil and observatory data. These findings provide important insights into the stability of Cs-GFO in the forest ecosystem. Considering the potential hotspot where Cs-GFO can accumulate deeper in the soil (>30 cm in depth), most Cs-GFO has remained in the forest for decades. Our study offers microscale heterogeneous Cs-GFO distribution in forests for ensuring long-term forest management planning necessary for both the long-term migration and local accumulation of Cs in forests.