Although the Forest Loss indicator in the 2014 EPI is a step toward better understanding changes in forest cover over the last decade, it is still far from perfect. Hypothetically, what would a set of ideal indicators for forests look like?
As with any issue, ideal forest indicators would be broadly measurable, reflective of performance, and relevant to policymakers. An ideal national-level measure would be in some ways scalable to local forests while encompassing the wide range of forest management objectives, from biodiversity, to carbon sequestration, to cultural and economic uses, and beyond. Another ideal measure would assess differences in forest species composition and richness – both prerequisites for sustaining a range of ecosystem services.1 Ideal indicators would also recognize that forest ecosystems are not isolated from other ecosystems but connected to them in form and function.
The use of satellite-derived data from Hansen et. al. (2013) to create a first-of-its-kind, high-resolution global map of forest extent, loss, and gain represents a step in the right direction. As mentioned in the Issue Profile: Forests, data for forest growing stock and forest cover change in past EPIs came from the UN Food and Agriculture Organization’s Forest Resource Assessment (FRA), which is primarily informed by country-reported data and is known to be plagued with inconsistencies across countries.2 Such disparities are exacerbated by various country definitions of what constitutes “forested land,” which in many cases is primarily determined by land use, or the ways forested areas are used by people.
A forest land-use definition might ignore certain types of disturbances that an analysis of forest cover might reveal. For instance, North American subtropical forests, which experience short-cycle planting and harvesting for lumber, underwent disturbance at nearly four times the rate of South American rain forests between 2000 and 2012. During that time, more than 31 percent of forest cover in the southeastern United States was lost or regrown. The resulting biophysical changes in form and function of those forests are ignored by the FRA’s land-use analysis. However, satellite data were able to capture it.3
A similar situation arises with respect to deforestation. A land-use methodology has significant implications for the measurement of deforestation. Huge changes in forest cover can be ignored if the reporting country does not list changes in use. This discrepancy was noted in the divergence in data between Canada and Indonesia, both of which clear natural forests without converting the land to non-forest land uses. However, because cleared forest that is left for “natural recovery” or listed as “managed forest” does not count as deforestation under the FRA’s guidelines, Canada’s loss of forest cover can go uncounted, while Indonesia’s is tallied.4
However, the Hansen et. al. (2013) analysis is not immune to definitional challenges, either. Because it tracks tree cover and is not yet fine-grained enough to differentiate between types of tree cover or the biophysical function of those trees, the Hansen et. al (2013) data cannot differentiate between reforestation—or the replanting of forest land—and afforestation, which may in fact have no direct biophysical function. This means Malaysia, with its numerous palm oil plantations, still receives credit for afforestation, even if these plantations are planted in grasslands, or for reforestation, even if those plantations replace natural growth forests. The Hansen et. al (2013) satellite-based mapping system of global forest change would benefit from differentiating between forest use practices to properly measure global forest change. It is possible, however, that no satellite will ever be able to fully capture such practical, grounded realities.
The EPI’s assessment of forest management would also benefit from the inclusion of other governance-based measures, such as whether countries make efforts to ensure the sustainable harvest and sale of timber and wood products (see Box: Measures of Stewardship – Forest Certification Schemes). Nonetheless, globally available data to assess more qualitative aspects of forest performance are sparse.
1 Chapin III, F. S., Zavaleta, E. S., Eviner, V. T., et al. (2000) Consequences of changing biodiversity. Nature, 405:234-242.
2 The Food and Agriculture Organization of the United Nations. (2008) Global Forest Resources Assessment 2010: Guidelines for country reporting for 2010. Available: http://www.fao.org/forestry/14097-0a8d9580c21c82a81175207120544e2b8.pdf. Last accessed: January 13, 2014.
3 Hansen, M., Potapov, P. V., Moore, R., et. al. (2013) High-resolution global maps of 21st-century forest cover change. Scienc, 342:850-853.
4 Hansen, M., Potapov, P. V., Moore, R., et. al. (2013) Supplementary Materials for High-Resolution Global Maps of 21st-Century Forest Cover Change. Science, 342:850-853.