Forest Landscape Integrity Index

• 1. Introduction

  ›

  The Forest Landscape Integrity Index arrived in Nature Communications on the 8th of December 2020. The study carried a number that surprised many forest scientists. 91.2 percent of the world's remaining forests had already been shaped, in some measurable way, by human activity. Each roughly 300-metre patch of forest on Earth now had a score between 0 and 10. A score of zero described land as thoroughly modified as it gets. A score of ten marked a forest that had essentially escaped human pressure. The research team was led by the Wildlife Conservation Society and drew in scientists from institutions across five continents. The questions it raised about the gap between designated protection and actual forest condition have since found their way into major international policy frameworks.

  1216

• 2. Mapping Human Pressure

  ›

  Grantham et al. built the index from four main spatial inputs, all processed in Google Earth Engine. The first was a map of forest extent as it stood at the start of 2019. Forest was defined as woody vegetation taller than five metres with at least twenty percent canopy cover. To establish which land counted as forest, the authors started from global tree-cover estimates for the year 2000. They then subtracted all mapped tree-cover loss from 2001 through 2018, excluding temporary canopy loss.

  The second and third inputs captured human pressure in two distinct forms. Observed pressures were activities that could be mapped directly at global scale: built infrastructure, agriculture, and recent deforestation. Inferred pressures addressed impacts that tend to radiate outward from roads, farms, and settlements but are harder to map directly. Selective logging, fuelwood collection, and hunting fall into this category. The model treated these inferred pressures as a decay function tied to proximity to observed pressure sources and access networks.

  The fourth input measured reductions in forest connectivity caused by loss and fragmentation. A forest pixel surrounded by cleared land received a lower connectivity score than an equivalent pixel embedded in a large unbroken block. For practical communication, the authors also produced a three-class map. Low integrity covered scores at or below 6.0, high integrity meant 9.6 and above, and medium fell between those thresholds. The class boundaries were benchmarked against reference sites of known ecological integrity, not set arbitrarily. The study noted the index could potentially be updated on an annual basis as new global datasets become available.

  161617

• 3. A Fragmented Planet

  ›

  In the 2019 map, 40.5 percent of the world's forest area qualified as high integrity, with a score of 9.6 or above. That amounts to roughly 17.4 million square kilometres. Medium-integrity forests covered another 33.9 percent, or about 14.6 million square kilometres. The remaining 25.6 percent, around 11 million square kilometres, fell into the low-integrity class with scores at or below 6.0. The global mean score across all forest pixels was 7.76.

  Russia and Canada alone contained roughly half of all high-integrity forest area in the 2019 map. Large tropical blocks also ranked highly: the Amazon, Central Africa, and New Guinea each contained substantial areas at or above the 9.6 threshold. Among the 172 countries ranked, 18 recorded national mean scores above 9.

  31.2 percent of the world's remaining forests were affected by directly observed human pressures. No single biome and no biogeographic realm had more than half its forest area classified as high integrity. What the scores revealed about the relationship between high-integrity forest and designated protected areas would prove among the index's most discussed findings.

  162

• 4. The Protected Area Gap

  ›

  Only about 27 percent of the world's high-integrity forest area fell inside a nationally designated protected area in 2019. That finding attracted immediate attention from conservation biologists. It meant that nearly three-quarters of the planet's most intact forest existed outside the boundaries of formal protection. The relationship looked different when viewed from the protected-area side: 56 percent of forests within designated protected areas qualified as high integrity. Protected reserves were, on average, better forest than the global baseline. But the vast majority of high-integrity forest lay beyond their reach.

  Those two numbers tell different stories about the same underlying gap. A reserve can contain excellent forest while still covering only a fraction of the high-integrity forest that exists across a landscape. The FLII made that distinction visible in a quantified and globally consistent form. An author correction published in 2021 revised an error in a protected-area table from the original article. It was a signal of how closely the research community was watching these particular figures.

  1186

• 5. Policy and Research Reach

  ›

  Target 2 of the Kunming-Montreal Global Biodiversity Framework lists FLII as a complementary indicator for monitoring ecological integrity, connectivity, and ecosystem restoration. The World Resources Institute's Global Forest Review uses the index as a measure of forest degradation in its synthesis of global forest change and condition. The 2023 Forest Declaration Assessment included "FLII units lost per year" as an indicator of forest degradation for tracking progress toward international forest goals. The European Commission's Joint Research Centre has embedded a Forest Landscape Integrity layer in its Global Forest Types 2020 map product.

  Sze et al. (2022) applied FLII in a pan-tropical study of forest integrity across the overlap of protected areas and Indigenous peoples' lands. Crowe et al. (2023) assessed forest integrity within thousands of Key Biodiversity Areas. BirdLife International has used FLII-based analyses to track trends in forest condition at sites identified for forest species. The index has also been used as an input for composite integrity metrics that draw on multiple global datasets.

  In conservation finance, the High Integrity Forest methodology, known as HIFOR, incorporates FLII thresholds in its eligibility criteria. Areas must meet those thresholds to be designated as high-integrity forest within the HIFOR monitoring framework. The constraints built into the underlying score matter for every one of these applications, and the authors were explicit about them from the start.

  437910

• 6. What the Index Cannot See

  ›

  The original paper described FLII as a conservative estimate of forest modification. That framing acknowledged two main sources of underestimation. Some human activities are genuinely difficult to map consistently at global scale. Fine-scale infrastructure and small-scale extraction may not register with the same reliability across all regions. Forest modification before the year 2000 falls largely outside what the underlying datasets capture. That means the index may not fully reflect historical cumulative impacts.

  Two additional drivers of forest change are not explicitly represented in the model at all: climate change and invasive species. Both can substantially alter forest structure, species composition, and ecological function. The forest extent layer introduces a separate complication. It can include tree crops and plantations in some cases. Those land covers can meet the canopy height and cover thresholds used to define forest. Where that happens, those areas tend to score as low integrity, so the effect is conservative rather than inflating the apparent picture of forest health. That gap between what the model counts as forest and what ecologists recognise as natural woodland is a limitation field studies have begun to address.

  1131416