Methodology

Current forest cover:

WRI used the only existing global map of current forest cover – that produced by the World Conservation Monitoring Centre (WCMC) in 1996, in collaboration with the World Wildlife Fund and CIFOR – as a base coverage for this assessment. The sources for this map include many country and regional maps derived from national and international sources, mostly accurate to a 1:1,000,000 scale. (All coverages used in the Frontier Forest Assessment were degraded to a 1:8,000,000 scale.) The quality, accuracy and dates of these national and regional maps vary. For details, please refer to: The World Conservation Monitoring Centre, The World Forest Map, (WCMC, Cambridge, 1996).

Mapping of Frontier Forest areas:

The frontier forest maps within this report reflect the knowledge and input of more than 90 experts and reviewers from around the world. Work on the Last Frontier Forests assessment began in May 1996, when WRI hosted meetings and review sessions (in Washington, Cambridge, U.K, and Rome) at which outside experts were asked to help define frontier areas and propose a methodology for mapping them. Mapping itself was a three-step process:

WRI first developed regional maps of “candidate frontier areas” by overlaying – using a Geographic information System (GIS) – the World Conservation Monitoring Centre’s current closed forest cover map with Sierra Club’s “wilderness areas” map [79] to define large wooded blocks devoid of human infrastructure (roads, settlements, etc.) where near-pristine forest might be found;

2. Approximately 90 regional forest experts were then asked to annotate these maps. Ten to fifteen individuals per region with broad knowledge of forest extent and condition reviewed the maps of candidate frontier areas and, using a checklist survey, either: (i) nominated candidate sites as “frontier forests”; (ii) redefined the boundaries of candidate sites or (iii) rejected candidate sites, and;

3. WRI digitized the revised boundaries of these areas.

Experts were also asked to fill in a questionnaire for each frontier area assessed. This survey was used to build a database for each frontier site, which includes information on:

i. Forest type (based on survey results). Experts were asked to classify frontiers according to one or more of the following categories: “boreal,” “temperate hardwood,” “temperate mixed coniferous and deciduous forest,” “southern temperate/astral forest,” “humid tropical forest,” “dry tropical forest.”

ii. Size (this was calculated using a GIS). WRI also asked experts to roughly estimate the size of each frontier assessed, where known.

iii. Threats to ecosystem integrity (based on information contained within the questionnaires). Experts were asked to rank sites as under high, medium, or low threat from “commercial logging,” “other biomass harvest (removal of fuelwood and construction materials, grazing),” “forest clearing (for agriculture, residential housing etc),” “road construction and other infrastructure development (e.g. powerlines, pipelines),” and “other,” and to provide additional details on known threats.

iv. Potential threats to ecosystem integrity (also based on information from the questionnaires). Experts were asked to: (i) note where significant portions of a frontier site have been allocated for current or future harvest (e.g. zoned as production forest, given over to concessions) and to estimate the percentage of each site allocated for harvest, where known; and (ii) indicate if a site contains high-value timber species, or other high-value resources (e.g. gold, oil), and to estimate the approximate proportion of the site within which such resources occur.

v. Source of expert information and level of knowledge (for most sites, based on the input of several reviewers). The names of these individuals are listed within WRI’s database, along with a ranking of how well these experts claim to know each frontier area (experts were asked to rank themselves on a 1-5 scale according to how well they know a given site).

In many cases, sites were nominated and reviewed independently by several experts. WRI harmonized these results by: (i) consulting with reviewers when their results differed for a given site to come up with a consensus view; (ii) in some cases, giving extra weight to results from individuals who ranked their knowledge of a site as significantly higher than that of others who reviewed that frontier forest.

WRI also held three review sessions (one covering temperate and boreal regions, a second covering tropical frontiers, and a third to review all areas of the world) to help fill in data gaps and to solicit input on draft frontier maps. Approximately 60 ecologists, foresters, and others with broad knowledge of forest issues across various regions of the globe attended these sessions.

Note that five frontier forest sites within Western Zaire were identified using maps of current road and forest cover (the latter derived from recent satellite imagery). These sites – lacking roads, or significant areas of degraded forest – were added because experts working on the WRI study suggested forest in this region qualified as “frontier,” however they were not sufficiently familiar with the area to identify the boundaries of these sites.

Mapping of original forest cover

As part of this project, WCMC developed the first detailed map of estimated forest cover “prior to the impact of modern man” (circa 8,000 years ago), using many global and regional biogeographic maps. These included World Map of Present Day Landscapes (Moscow State University/UNEP), Ecoregions of Latin America and the Caribbean (Dinerstein et al.), Vegetation of Africa (White), Review of the Indomalayan Protected Areas (MacKinnon), Australia Natural Vegetation (Australian Surveying and Land Information Group), and General Map of Natural Vegetation for Europe (Bohn and Katenina). WRI adjusted this map – notably by using current forest cover in the northernmost 200 km of forest as a surrogate for historic cover so as to capture the natural patchy distribution of forest in this tundra-forest transition zone.

The WCMC map is an indicator, not a direct measure of original cover. It depicts where forest might be expected to occur today in the absence of humans, based on climate, topography, and other variables. The distribution of forest cover 8,000 years ago probably varied somewhat from region to region, due to long-term climate change. For details on the original forest cover mapping methodology, please refer to:

Clare Billington, Valerie Kapos, Mary Edwards, Simon Blyth and Susan Iremonger, Estimated Original Forest Cover Map A First Attempt (WCMC, Cambridge, UK, October 1996).

Frontier Forest Index:

The frontier forest index ranks countries on a scale of 0 to 99. The index was created by multiplying the percentage of original frontier forest lost (that estimated to cover the country 8,000 years ago) with the percentage of remaining frontier forest area classified under moderate or high threat to total assessed frontier area.

References and notes

79. Michael McCloskey and Heather Spalding, “A Reconnaissance-level Inventory of the Amount of Wilderness Remaining in the World,” Ambio, Vol. 18, No. 4, 1989, pp. 221-27.