Humboldt-Universität zu Berlin, Agricultural Economics and Social Sciences, Germany
Soil degradation increasingly threatens food supply in many Southeast Asian countries. Socio-economic factors are considered crucial for the state of soils. Yet the understanding of links between socio-economic factors and soil degradation remains limited, with competing hypotheses existing about links to the most often cited causes of soil degradation: population pressure and poverty. At the same time data sources are limited, but allow for analysing the problem at a regional level. This paper contributes to the discussion by conducting a statistical analysis at a regional level. The analysis is based on a pressure-state-response framework, in which the influence of causal factors of soil degradation depends upon the framework conditions. This allows for varying links in different systems with similar natural and socio-economic conditions. The profiles of framework conditions are conceptualised as agricultural development patterns. The soil degradation data considered is the qualitative "Assessment of the Status of Human-Induced Soil Degradation in South and Southeast Asia" on which technical data improvements and limited validity checks have been conducted. Further geo"=referenced data sets available for the region are overlaid in a Geographic Information System (GIS). Poverty data have been assembled from both national Poverty Assessments and Human Development Reports.
Exploratory methods (factor, correspondence, and cluster analysis) were used to empirically structure context factors in form of agricultural development patterns. A subsequent analysis provided information on the incidence of different forms of soil degradation within clusters and on the influence of different factors on water erosion.
Results show, that by including socio-economic variables in regression analysis, explanation of water erosion can be improved as opposed to regressions including only natural conditions, showing a positive relation of poverty and population density with water erosion for the region as a whole. Clusters reveal clear differences with regard to the incidence of soil degradation. Regression analyses show differing importance of causal factors within clusters.
The approach proves useful to provide better knowledge about critical constellations of natural, socio-economic and land use factors with regard to soil degradation. Thus it enhances prospects for geographic targeting of measures against soil degradation.
Keywords: Agricultural development, GIS, Multivariate analysis, soil degradation