Conference Agenda

Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).

 
Session Overview
Session
314R: Forest transitions and the resurgence of tree cover in the Global South
Time:
Friday, 26/Apr/2019:
3:00pm - 4:15pm

Session Chair: Jefferson Fox
Session Chair: Kaspar Hurni
Location: MB-114
Main Building, 1st floor, west wing, 78 seats
Session Topics:
How do we support transformation?

Session Abstract

Mather (1990) described large-scale forest regrowth, reforestation, and afforestation as the ‘forest transition’, a trajectory of change where initial forest loss is followed by recovery as a country undergoes social and economic changes. Hecht (2010) criticized prevailing forest-transition models for failing to examine the effect of globalization on forest cover, suggesting that the globalization of labor, discourses, knowledge, capital and new emergent markets provide a better optic for understanding the complexity of forest recoveries and that forest-transition theory does not currently capture these forces.

This session will explore forest transition and the resurgence of forest and tree cover in the Global South. It is important to distinguish between changes in tree cover and ‘natural forest’ because the loss of natural forest may go unnoticed if plantation forests displace natural forests in the course of forest transition. Scholars have hypothesized a number of factors as affecting tree cover [i.e., such as an active community forest management program (Nepal) or engagement in the global labor market, a key link in the deagrarianization process (the Philippines)]. The concurrent impact of community forestry programs and the delinking of household livelihood and land resources (both agriculture and forest) would presumably result in better extent and quality of forest and tree cover due to a changing agricultural landscape. This session invites seek papers addressing such issues of forest transitions and tree plantations, telecoupling, REDD, and forest definitions from across the Global South.

Session Organizers: Jefferson Fox and Kaspar Hurni


Presentations
Full talk
ID: 900 / 314R: 1
314R Forest transitions and the resurgence of tree cover in the Global South
Keywords: forest cover mapping, forest transition, Nepal. LandTrendr

The creation of annual forest cover maps to assess forest transition in Nepal from 1990-2016

Kaspar Hurni1,2, Jamon Van Den Hock3, Alexander Smith3

1East West Center, HI, USA; 2Center for Development and Environment (CDE), University of Bern, Switzerland; 3Geography and Geospatial Science, Oregon State University

We created an annual time-series of binary forest and non-forest maps from 1990-2016 in Nepal using the random forest classifier and a rigorous pre-processing of Landsat images to reduce data gaps and noise related to cloud cover and haze, inter- and intra-annual variations in vegetation greenness, and topography. We based our analysis on Landsat 5, 7, and 8 surface reflectance data, which we first corrected for differences in the illumination condition related to topography. Next, we performed an image composition using growing season images (July-October) of each year to obtain a time-series of annual image composites. To further eliminate data gaps we then evaluated land cover trends for each pixel from 1992-2016 using LandTrendr. This allowed us to estimate the land cover reflectance for those years and pixels where data gaps occurred, ultimately resulting in a gap- and noise-free annual time-series of Landsat data. We then collected forest and non-forest classification samples using the 1990 and 2010 imagery. Due to our rigorous image pre-processing routine, which normalized the data across time and space, we could merge the samples collected for these two years and apply them to each year of the time-series to obtain annual forest cover maps. This approach proved to work well and we reached a 90% overall classification accuracy with 87% forest user’s accuracy. We found nearly 19% tree-cover expansion over the study period and that the relative gains of tree cover were comparable in the Middle Hills and Mountain regions of Nepal while we mapped less gain in the Terai. We also examined tree cover change at the more local Village Development Committee (VDC) level as well, identifying that the most gain has occurred in approximately twenty VDCs spanning central and eastern Nepal.



Full talk
ID: 899 / 314R: 2
314R Forest transitions and the resurgence of tree cover in the Global South
Keywords: afforestation, Nepal, LCLUC, machine learning

Mapping changes in tree cover in Nepal 1990 to 2016: Understanding the drivers of change

Jefferson Fox, Sumeet Saksena, Hanpei Zhang

East-West Center, United States of America

We developed statistical models to explain tree cover using a large suite of biophysical and socio-economic data. Considering that most coupled natural-human systems are complex, the team felt it best to use complimentary modeling approaches. Specifically we chose to use multi-level regression with Village Development Committees (VDC), the smallest administrative unit in Nepal, nested under districts (aka mixed effects or hierarchical modeling) and a machine learning method – Random Forests (RF). In both these models, we also incorporated the effects of spatial clustering. Initially we chose to model as the dependent variable the change in percentage tree cover between 2001 and 2016. The rationale for this is that any socio-economic changes that happened around 2001 (the census year) could have possible impacts on the tree cover shortly after 2001. We found that places with less tree cover, to begin with, had higher growth rates. We also found the number of migration to be significant only in the Middle Hills. The positive impact of community forestry was observed also mainly in the Middle Hills, as was the effect of distance to the nearest protected area. The Random Forest Model suggested that population density is one of the key drivers of change, more so than either migration or community forestry or protection efforts. An advantage of the Random Forest models over the multi-level models is that they enable us to examine the partial dependence between the outcome variable and the predictors. In all cases, we gained valuable insights into these relationships. They are nonlinear, non-monotonous, and they have thresholds, tipping points, and saturation effects. Previous studies of the impact of migration on tree-cover change did not explore these aspects of the association. In addition, Fox and country collaborator, Dr. Ram Chhetri, have conducted fieldwork in six Nepali villages since the 1980s. An examination of changes in agriculture, livelihoods, and tree cover in these villages provides a finer lens through which to interpret the modeling results.



Full talk
ID: 339 / 314R: 3
314R Forest transitions and the resurgence of tree cover in the Global South
Keywords: swidden, agricultural intensification, land-sparing, rural policy, Mexico

The influence of modern state institutions on forest transitions: The case of the Southern Yucatan, Mexico

Carlos Dobler-Morales1, Rinku Roy Chowdhury1, Birgit Schmook2

1Clark University, United States of America; 2ECOSUR-Chetumal, Mexico

By the 1980s, the Southern Yucatan, in southeast Mexico, had consolidated its status as a global hotspot of deforestation. Forest conversion took place as a consequence of rapid smallholder occupation, a process accompanied by swidden agriculture expansion. Today, however, the region exhibits considerably lower forest loss rates. This paper examines the factors that influenced this transition. In particular, this study focuses on the role that modern agrarian, welfare, and environmental policy has played in changing land-use patterns in the region. As elsewhere in Latin America, smallholder livelihoods in the Southern Yucatan have not only become more globally connected through distant labor opportunities and remittances; they have also become exposed to a panoply of agricultural subsidies and antipoverty programs, as well as expanding protected areas and new conservation instruments such as Payment for Ecosystem Services. Drawing from an 81-household survey, 17 key-informant interviews, and a 31 year Landsat time-series, this study traces the linkages between the region’s shifting policy environment, smallholder land-use decisions, and their landscape-level expressions. Results illustrate how the array of state institutions operating in the Southern Yucatan have been inducing smallholders to intensify their agricultural production via incentives to “modernize” their swiddens and restrictions on rotating fields. At a landscape level, these changes in agricultural practices appear to be contributing to the stabilization and simplification of the formerly dynamic and complex agro-forest mosaic. The case of the Southern Yucatan thus represents a trajectory of forest recovery that does not fit neatly with interpretations rooted in the deagrarianization of the countryside. Instead, agricultural livelihoods persist here, often expressing the ways smallholders navigate the institutional arrangement that simultaneously encourages them to intensify their production while setting apart land for conservation.



Full talk
ID: 287 / 314R: 4
314R Forest transitions and the resurgence of tree cover in the Global South
Keywords: deforestation, forest transition, agriculture, forestry, trade

Deforestation displaced: trade in forest-risk commodities and the prospects for a global forest transition

Florence Pendrill1, U. Martin Persson1, Javier Godar2, Thomas Kastner3

1Chalmers University of Technology, Sweden; 2Stockholm Environment Institute; 3Senckenberg Biodiversität und Klima

While many developed countries are increasing their forest cover, deforestation is still rife in the tropics and subtropics. With international trade in forest-risk commodities on the rise, it is becoming increasingly important to consider between-country trade linkages in assessing the drivers of—and possible connections between—forest loss and gain across countries. Previous studies have shown that countries that have undergone a forest transition (and are now increasing their forest cover) tend to displace land use outside their borders. However, lack of comprehensive data on deforestation drivers imply that it has not been possible to ascertain whether this has accelerated forest loss in sourcing countries. To remedy this, we present a land-balance model that quantifies deforestation embodied in production of agricultural and forestry commodities at country level across the tropics and subtropics, subsequently tracing embodied deforestation to countries of apparent consumption using a physical, country-to-country trade model. We find that in the period 2005–2013, nearly 60% (5.3 Mha yr-1) of forest loss could be attributed to expanding commercial cropland, pastures and tree plantations. The commodity groups most commonly associated with deforestation were cattle meat, forestry products, oil palm, cereals and soybeans, though variation between countries and regions was large. A large (26%) and slightly increasing share of deforestation was attributed to international demand, the bulk of which (87%) was exported to countries that either exhibit decreasing deforestation rates or increasing forest cover (late- or post-forest transition countries), particularly in Europe and Asia (China, India, and Russia). About a third of the net forest gains in post-forest transition countries was in this way offset by imports of commodities causing deforestation elsewhere, suggesting that achieving a global forest transition will be substantially more challenging than achieving national or regional ones.