Main content area

Identification and quantification of drivers of forest degradation in tropical dry forests: A case study in Western Mexico

Morales-Barquero, Lucia, Borrego, Armonia, Skutsch, Margaret, Kleinn, Christoph, Healey, John R.
Land use policy 2015 v.49 pp. 296-309
basins, case studies, correlation, dry forests, forest resources, interviews, land use, landowners, landscapes, livestock, population size, probability, regression analysis, shifting cultivation, socioeconomic factors, surveys, sustainable forestry, Mexico
The intensity of forest degradation is linked to landowners’ decisions on management of their shifting cultivation systems. Understanding the processes involved in this land use type is therefore essential for the design of sustainable forest management practices. However, knowledge of the processes and patterns of forest transition that result from this practice is extremely limited. In this study, we used spatially-explicit binary logistic regression to study the proximate factors that relate to forest degradation by combining biophysical and socio-economic variables. Our study region is within the Ayuquila Basin, in Western Mexico, a typical fragmented tropical dry forest landscape dominated by shifting cultivation. Through a survey and semi-structured interviews with community leaders, we obtained data on the forest resources and on the uses that people make of them. Detailed forest cover maps for 2004 and 2010 were produced from high-resolution SPOT 5 data, and ancillary geographical data were used to extract spatial variables. The degree of social marginalization of each community and the ratio of forest area to population size were the main factors positively correlated with the probability of the occurrence of forest degradation. Livestock management and use of fence posts by the communities were also positively associated with forest degradation. Among biophysical factors, forest degradation is more likely to occur in flatter areas. We conclude that local drivers of forest degradation include both socioeconomic and physical variables and that both of these factors need to be addressed at the landscape level while developing measures for activities related to REDD+.