Jeffrey Chow, PhD

2012 TRI Fellow in El Salvador

Reforestation Dynamics of Agricultural Lands in El Salvador


The concept of the forest transition describes the tendency for forest cover to decrease in response to colonization and population growth, and then subsequently rebound as societies undergo economic development, industrialization and urbanization.  Spatially explicit regression models measure the impact on land cover of geographic characteristics while accounting for location relative to human settlements, markets, and transportation networks.  With the availability of remotely sensed data and geographic information systems, these models have often have been used to explain patterns of clearing in natural forests .  In contrast to clearing, reforestation has been less frequently studied using these techniques, and the characteristics that encourage reforestation are less well understood.  This study asks: what local ecological, geophysical, and socioeconomic characteristics have encouraged forest resurgence on former cultivation and pasture lands in El Salvador?  

The analytical framework employed in this study relies on the assumption that land use decisions are made by profit-maximizing private landowners.  El Salvador provides a fitting case study because land exists predominantly under private ownership, either by individuals or cooperatives.  The vast majority of privately owned lands, including forested areas, are subject to secure tenure.  Exceptions to this condition include areas which are designated as protected areas but not yet legally transferred to government ownership.

This research is part of a dissertation study using geophysical, socioeconomic, and remotely-sensed land cover data to econometrically determine which characteristics have yielded reforestation in El Salvador.  Ground-truthed data collection was undertaken to improve the accuracy of remotely-sensed land cover models to be input into the econometric analyses.  Sites were preselected based on a stratified random sample from fifteen land cover change classes determined from a preliminary classification using Landsat5-TM data.   Site visits and conversations with landowners and land managers were implemented to confirm shifts in vegetation types between the early 1990s and today.