Sulphur (S) is one of the essential elements for both higher plants and animals, as it is known to be a constituent of S-containing amino acids such cysteine and methionine and hence of proteins. However, S deficiency for crops has become increasingly widespread worldwide, probably due to decreased inputs from the atmosphere and fertilizers. In this context, comprehensive information on the S status in soil would be inevitable to prevent S deficiency for plants and for animals. The objectives of this research were (1) to evaluate the total and available S contents and (2) to elucidate their determining factors of agricultural soils in Japan.

One hundred and seventy-eighty soil samples with variable soil types were collected from the surface layer (0-15 cm) of agricultural fields all over Japan. Ninety-five samples were collected from paddy and 83 samples were collected from upland. Total S content was determined by the energy dispersive X-ray fluorescence analysis (EDXRF) using 2-4 g of pressed soil pellets. For the determination of available S, air-dried soil was extracted with 0.016 mol L^-1 KH₂PO₄ solution for 30 min at a soil to solution ratio of 1:5 and the S concentration of the extract was determined by ICP-AES. General physicochemical properties of the soil samples including pH, total C and amorphous Fe and Al contents (Fe_o, Al_o) were also investigated.

The total S content ranged from 0.11 to 11.8 g kg^-1 with median and arithmetic means of 0.91 and 1.20 mg kg^-1, respectively. The median of total S was slightly higher than the reported value for world soils, i.e. 0.7 mg kg^-1. Two samples from Hokkaido showed extremely high total S content, possibly due to volcanic activity. The available S content ranged from 4.5 to 662 mg kg⁻¹ with median and arithmetic means of 50.7 and 81.3 mg kg⁻¹, respectively, and corresponded to 6.8 % of the total S on average. According to the threshold of the available S, only 4 samples out of 178 were regarded as S deficient. Correlation analysis for 176 samples except for 2 outliers showed that total S content had a significant positive correlation with the total C, Al_o, Fe_o, Fe_d, Fe_o and Al_o + 1/2Fe_o contents (p < 0.01). This result strongly suggests that organic matter in combination of amorphous Fe and Al oxides were the main factors determining the total S content. The available S content had a significant positive correlation with Fe_d, Al_o, Fe_o, Fe_d-Fe_o, Al_o + 1/2Fe_o and total C contents as well as total S content and had a significant negative correlation with sand content (p < 0.01), suggesting the contribution of amorphous and crystalline Fe and Al oxides in addition to organic matter. They also showed significant positive correlation with total and available Se contents, suggesting high similarity between S and Se. Among soil group, red soils (Udults) and volcanic soils (Udands) showed relatively high total and available S contents, due to their abundance of crystalline and amorphous Fe and Al contents, respectively. In conclusion, agricultural soils in Japan were relatively abundant of S mainly due to high amorphous minerals content reflecting volcanic activities. These findings can be used as basic information on soil S status in agricultural soils, on which sustainable agriculture and environmental conservation are to be established, not only in Japan but also in other countries with the influence of volcanic activities.