Abstracts of Nippon Dojo-Hiryogaku Zasshi 93 - 2

Based on the fact that humus is dark and stable against micro-bial degradation, we hypothesized that stable humus content can be estimated by the soil carbon content predicted from spectral reflectance. Furthermore, the difference between the total carbon content and the predicted value can be attributed to easily-decomposable organic matter, representing available nitrogen. Thus, we suggest a new method to estimate available nitrogen based on total carbon content and spectral reflectance. To test this method, we conducted drone-based remote sensing using a four-wavelength multispectral camera at paddies, where livestock man-ure compost and rice straw has been continuously applied, and paddies where no organic matter has been added after puddling. Using spectral reflectance, the model tended to underestimate the predicted carbon content in such paddies. Coefficients of determination were higher in paddies without added organic matter, suggesting that the value predicted via spectral reflectance may measure stable humus. We found a significant correlation between the available nitrogen and the difference between the predicted value from the spectral reflectance and the total carbon content. This finding indicates that total carbon content and spectral reflectance can be used to estimate available nitrogen. Using both the spectral reflectance and the total carbon content in the predictive formula, the coefficient of determination increased from 0.70 to 0.90 in air-dried soil incubated for 4 weeks and from 0.45 to 0.74 in wet soil incubated for 10 weeks. The applicability of this method should be verified in different soil and humus types.