Large-scale land investments and their socio-environmental effects

Kyle Frankel Davis and Saleem H. Ali, University of Delaware, Jacopo Baggio, School of Politics, Security, and Internaional Affairs, University of Central Florida, Orlando, FL, Paolo D'Odorico, Department of Environmental Science, Policy, and Management, University of California at Berkeley, Berkeley, CA, Jampel Dell'Angelo, VU University Amsterdam, Sandra Eckert, University of Bern, Lyndon Estes, Graduate School of Geography, Clark University, Worcester, MA, Suhyun Jung, West Virginia University, Laura Kehoe, Oxford University, Milad Kharratzadeh and Heejin Irene Koo, Columbia University, Tobias Kuemmerle, Humboldt University Berlin, Berlin, Germany, Domingos Machava, Universidade Eduardo Mundlane, Marc Muller, University of Notre Dame, Aurélio de Jesus Rodrigues Pais, Universidade Zambeze, Natasha S. Ribeiro, Forest Engineering, Eduardo Mondlane University, Maputo, Mozambique, Maria Cristina Rulli, Politecnico di Milano, Mokganedi Tatlhego, University of California Berkeley

Background/Question/Methods
Forests are vital for global biodiversity, carbon storage, and local livelihoods, yet ongoing globalization has meant that remote factors exercise growing influence over these forests and their common-pool resources. Large-scale land acquisitions (LSLAs) by foreign investors have emerged as an important mechanism linking the demands of distant actors to forests in the Global South. Yet it remains unclear whether LSLAs preferentially target forested areas, whether they have influenced rates of forest loss, and what their future implications may be for biodiversity in targeted areas. Here we employ a mixed methods approach to perform a global assessment of the effects of LSLAs on forests - combining georeferenced maps of individual land deals for tree plantations, mining, and logging, geolocated information of agricultural LSLAs including contract year and intended use, and data on annual forest cover.

Results/Conclusions
Compared to non-investment areas, we find that LSLAs were granted in areas of higher forest cover in 81% of cases, contributed disproportionately to forest loss in 52% of cases, and produced slight but significant decreases in forest loss in 39% of cases. The remaining cases (9%) saw no significant change in rates of forest loss. LSLAs for oil palm, tree plantations, and wood fiber were linked to greater forest loss while those for logging and mining showed a mix of outcomes. For agricultural LSLAs, we find a significant enhancement in rates of forest loss, with the potential to produce losses in relative species richness and mixed outcomes for relative abundance. These findings provide important context for the influence of LSLAs on tropical forests.