................................................................................................................................. 02 Acknowledgements............................................................................................................... 03 Introduction............................................................................................................................ 05 Objectives............................................................................................................................... 09 Study Area.............................................................................................................................. 10 Methods.................................................................................................................................. 12 Data............................................................................................................................ 12 Landcover................................................................................................................... 13 Predictors................................................................................................................... 13 Statistical Model........................................................................................................ 15 Analysis...................................................................................................................... 16 Results.................................................................................................................................... 17 Landcover................................................................................................................... 17 Geospatial Analysis.................................................................................................... 19 Statistical Analysis..................................................................................................... 22 Deforestation Scenarios............................................................................................. 24 Discussion.............................................................................................................................. 29 References.............................................................................................................................. 32 Appendix................................................................................................................................ 35 Is the Inter-Oceanic Highway exporting Deforestation? César I. Delgado Masters Project Final Draft 5 of 37 Introduction The Inter-Oceanic highway is the final chapter of the infrastructure integration efforts between Peru and Brazil, joining the two coasts of the South American continent through a continuous paved road that bisects southern Amazonia, under the promotion of the Initiative for the Integration of Regional Infrastructure in South America (IIRSA). It is estimated that this road will benefit more than 13.3 million dwellers of the triple border region between Brazil, Bolivia and Peru (IIRSA 2007). However, the international scientific community has growing concerns about the possible impacts to the relatively pristine Peruvian and Bolivian tropical rainforests due to probable land-use changes triggered by the socioeconomic integration of these two countries with Brazil. These fears are based on the idea that the highway will not only spur the trade of goods and services between these countries, but that Brazils appalling deforestation rates will exacerbate the deforestation trends within Peru and Bolivia. On the other hand, supporters of the road strongly believe that road construction is an essential step to reduce poverty in these remote Amazon regions. Although the debate carries on, the road is scheduled to be fully operational by the year 2009, therefore the changes are already in motion and the uncertainties of its influence will start to be revealed in the near future. It is essential to characterize a deforestation baseline prior to the full completion of this road project. This baseline will be essential to quantify, if any, the deviation from the recent deforestation trends and reveal if the Inter-Oceanic highway is not only exporting goods and services, but deforestation too. 1 The Initiative was adopted at a meeting of South American presidents held in the city of Brasilia, Brazil, in August 2000, at which the region’s leaders agreed to take joint actions to promote South American political, social and economic integration that includes the streamlining of regional infrastructure and specific measures to foster the integration and development of isolated sub-regions. http://www.iirsa.org/ 2 According to 2000-2005 UN-FAO statistics Brazil has the highest average annual deforestation rate of primary forests with 3.4 million hectares per year. Is the Inter-Oceanic Highway exporting Deforestation? César I. Delgado Masters Project Final Draft 6 of 37 According to some scientists, deforestation models on South America present certain bias towards treating the Amazon basin as a political and socioeconomic continuum, failing to acknowledge that the Amazon basin is roughly the size of the United States of America consisting of more than six countries. They argue that deforestation predictions must take into account national differences, including socioeconomic organization, population characteristics and dynamics, government systems, public policies, environmental laws and degree of enforcement, (Soares-Filho et al. 2006, Perz et al. 2005). Geist and Lambin (2001) separate tropical deforestation drivers in three proximate causes (agricultural expansion, wood extraction, expansion of infrastructure) and five broad categories of underlying driving forces (demographic, economic, technological, policy/institutional, and cultural or socio-political factors). Therefore, Figure 1: The Inter-Oceanic highway route and the location of the triple border between Brazil, Bolivia and Peru. (Original source: MTC-Peru) Triple border region Interoceanic highway: Paved Unpaved Is the Inter-Oceanic Highway exporting Deforestation? César I. Delgado Masters Project Final Draft 7 of 37 we cannot predict in a reliable manner the deforestation of a multinational road project such as the Inter-Oceanic highway by relying in current Pan-Amazonian deforestation models that do not account for the national differences in the underlying factors. Several empirical techniques are typically used to analyze the driving factors, the relationships between driving factors, and related decision making processes of land use change (Lesschen et al. 2005). The actual use of empirical techniques differs: often the prime interest of social scientists is explanation of observed land use changes, while ecologists focus on prediction (Lesschen et al. 2005). In this particular case I not only approach a land-use change model as a social scientist by including the secondary road network, the population centers (markets), and national deforestation dynamics due to cultural or socio-political factors, but also as an environmental scientist by predicting the impacts of deforestation. Finally, it is important to acknowledge in the scope of this project that the term deforestation exclusively refers to clearcutting of the forest. Therefore, selective logging and forest degradation by non-timber extractive activities are excluded. It is also important to understand that tropical rain forest is the predominant land-cover of the study area, and that any land-use change will consequently generate a lost in forest cover. Is the Inter-Oceanic Highway exporting Deforestation? César I. Delgado Masters Project Final Draft 8 of 37 Objective The primary objective of this project is to answer the question that entitles this research project: Is Inter-Oceanic highway exporting deforestation? In other words I will determine if Brazilian deforestation dynamics will exacerbate the ones in Bolivia and Peru, due to the pavement of the Inter-Oceanic highway and geospatially predict the possible deforestation scenarios on this region of the lowland Amazon rainforest. However, in order to reach this objective is essential to accomplish the secondary goals stated below. • Quantify the deforestation rates regionally and within each country in the study area for the study time periods (1989-2000). This deforestation geospatial analysis will not only depict the land-use changes on the region, but identify the key geospatial deforestation drivers. • Create a series of statistical models to detect the maximum likelihood estimator (MLE) of each geospatial deforestation driver within each country. Besides the fact that these models will allow the selection of significant geospatial deforestation drivers, they will create a temporal scale to better understand the relation between deforestation and human infrastructure. • Rely on the MLE’s to project the current deforestation trends into the near future. Two scenarios were created; a “Pan-Amazonian” under which there are no differences between the deforestation dynamics of the countries in the study area, and the “Controlling by Country” under which the projection will internalize the historic differences in the deforestation dynamics of the countries involved. Is the Inter-Oceanic Highway exporting Deforestation? César I. Delgado Masters Project Final Draft 9 of 37 Study Area The study area engulfs more than 13 million ha of the Low Tropical Wet Forest located between 110 and 520 m above sea level, on the triple border of the Amazon provinces of Madre de Dios (Peru), Acre (Brazil), and Pando (Bolivia). In ecological terms this is one of the most species-rich biomes in the world, harboring more than 1/3 of all species on the globe (Turner 2001). However, the importance of this region goes beyond its natural beauty and its genetic diversity, it is co