ドイ　リョウイチ   土居　良一   社会学部 社会学科   教授

■ 標題   Doi 2005. Human-induced land degradation gradient shown by antibiotic susceptibility profiles of bacterial communities and physico-chemical soil quality characteristics

■ 概要   In this study, multivariate profiles of soil bacterial communities were used to measure the intensity of land degradation as a result of deforestation of a Thai tropical forest and subsequent human activities. Soils were sampled at a dry evergreen forest (the original vegetative type), a dry deciduous forest (moderately disturbed) and bare ground (the most severely degraded). The soil bacterial communities were profiled by the antibiotic disk diffusion (ADD) method. Based on antibiotic susceptibility profiles of soils, the ADD method may give soil quality measures that are unique in their meanings. Principal component analysis extracted principal components from the antibiotic susceptibility and soil physico-chemical data sets. Both the first principal components showed the degradation gradient in the principal component score plots. Redundancy analysis showed that decreases in soil moisture, total carbon content and electrical conductivity involved in the degradation correlated significantly to changes in the antibiotic susceptibility profile. Multiple regression analyses between values of a soil fertility index (SFI) or a soil evaluation factor (SEF) and scores on the principal components or redundancy analysis ordination axes indicated that the first principal component or redundancy analysis ordination axis significantly explained the SFI and the SEF models (p<0.002). The formulae that quantified the intensity of the land degradation are given. This scheme is expected to help the conservation and rehabilitation of lands by predicting results, describing a gradient of interest and suggesting preferable soil conditions for particular goals.   単著   Nat Hum Act   9,pp.33-45   2005

Copyright(C) 2011 Daito Bunka University, All rights reserved.