Abstract
Tropical forests are disappearing at an alarming rate. In Central America, a hectare of forest is cleared for agriculture every 5 min. This study was conducted in a forested 4,000 ha watershed of central Honduras to find deforestation causes based on socio-economic characteristics of population. First, a multitemporal analysis of Landsat TM imagery was conducted to determine deforestation rates and agricultural–forest boundaries. A GIS buffer procedure allowed determining which households were at the deforestation front and which households were located at the rest of the area (control). GIS techniques were used to extract biophysical information such as slope, elevation, land cover, temperature, precipitation, etc. Then, we set up a data base with more than 50 socioeconomic variables (level of education, income, children per family, major economic activity, use of conservation practices, etc.). Around 500 households, distributed all over the watershed, were visited, interviewed and GPS-located. A multivariate statistical analysis allowed an exploratory analysis to eliminate non useful and redundant variables and then to determine what variables appear to be important predictors of deforestation behavior among rural families. A resulting logistic regression model showed that household with lower annual income heads and with less use of conservation practices were more statistically prone to clear the forest (α = 0.001). The study uncovered the complexity of this problem and confirmed the need of using GIS–remote sensing techniques to combine socioeconomic and environmental data in several time–space dimensions to find the causes and trends of tropical deforestation.
Full citation
Rivera S., Martinez de Anguita P., Ramsey R.D., Crowl T.A. (2013). Spatial Modeling of Tropical Deforestation Using Socioeconomic and Biophysical Data. Small-scale Forestry, 12:321-334. 10.1007/s11842-012- 9214-2.