How "Virgin" Is Virgin Rainforest?

See allHide authors and affiliations

Science  16 Apr 2004:
Vol. 304, Issue 5669, pp. 402-403
DOI: 10.1126/science.1093991

Conservation biologists increasingly use the term “wild nature” rather than “high biodiversity” to identify blocks of biodiverse habitats that have been relatively undisturbed by human activity (1). Their preference for this term is driven by frustration that vast swathes of biodiverse habitats continue to be lost at unprecedented rates while biologists argue about which “currency” is best for measuring the value of biodiversity—genetics, species, family, rarity, endemicity—and which regions should be selected for conservation efforts. This is especially true for the tropical rainforests which, according to current estimates, are disappearing at a rate of ∼6 million ha per year (2). Alongside these depressing rates of destruction, evidence has started to emerge from archaeological and paleoecological investigations that many of these so-called “virgin” rainforest blocks might not be as pristine as originally thought and have in fact undergone substantial prehistoric modification. The implications of such studies for understanding the resilience and recovery of tropical rainforests following human disturbance are far-reaching and should not be overlooked by conservation biologists.

The three largest undisturbed rainforest blocks are in the Amazon basin, lowland Congo basin, and the Indo-Malay region of Southeast Asia (see the figure). Yet a number of case studies in each of these regions now suggest that prehistoric human activities were far more extensive than originally thought. In the Amazon basin, for example, recent studies indicate that regions with the most fertile soils in the lowland rainforest are those with “terra preta” soils (3). Formation of these soils is attributed to prehistoric burning and agricultural activities from around 2500 years ago, and in central Amazonia, estimates suggest that terra preta soils cover up to 50,000 ha. In addition, emerging archaeological evidence from the Upper Xingu region of Brazil indicates extensive late prehistoric settlements dating between ∼1250 to ∼1600 A.D., covering regions up to 40 to 80 ha, and supporting populations between 6 and 12.5 persons per km2 (4). These were complex regional settlements indicating intensive management and development of the landscape and resulting in large-scale transformation of the forest to agricultural land and parkland. Interestingly, abandonment of the land following catastrophic depopulation between 1600 and 1700 A.D. resulted in extensive reforestation in many areas. The Upper Xingu region of Brazil now comprises the largest contiguous tract of tropical forest in the southern peripheries of the Amazon.

Anthropogenic modification of “virgin” tropical rainforest.

The map indicates the three largest undisturbed rainforest blocks remaining worldwide. These are located in the Amazon basin, the Congo basin, and Southeast Asia (Indo-Malay region). Evidence is accumulating from archaeological and paleoecological studies that each of these regions had been disturbed by prehistoric human settlements but had subsequently regenerated once the populations moved on or died out. This suggests that given sufficient time, tropical rainforests disturbed by modern human activities may be able to regenerate.


A combination of archaeological and paleoecological studies reveals a similar story in the lowland Congo basin. Here, there have been extensive finds of stone tools, oil palm nuts, charcoal horizons (subsoil layers of charcoal), banana phytoliths (silica bodies found in plants that are preserved in sediments and permit identification of the source plant), and pottery fragments (5, 6). These discoveries have led to the conclusion that much of this region underwent extensive habitation, clearance, and cultivation beginning ∼3000 years ago and ending ∼1600 years ago, following a population crash. In western central Africa there is also archaeological evidence for iron-working furnaces dating from ∼650 B.C.—another activity that would have had a serious impact on the forest through the extraction of wood for charcoal production and smelting. A population crash in the fifth century A.D. resulted in abandonment of the land and widespread forest regeneration throughout these regions (7). Many forest types resulting from this former human occupation are still to be found in the lowland Congo basin. In some areas, often considered “virgin,” the forests may still be undergoing a process of secondary succession (8).

There is even earlier evidence of prehistoric modification of the tropical rainforest in the Indo-Malay rainforest block. This disturbance has affected the composition and diversity of present-day species. Archaeological and archaeobotanical evidence suggest that agriculture (including the cultivation of bananas) was established ∼7000 years ago in Papua New Guinea (9) and that the lowland rainforests of Thailand have been managed from as early as 8000 years ago. In the rainforests of Thailand, human occupation and management of the land may have increased tree diversity (10). Finally, in the Solomon Islands, evidence from archaeology, ecology, and oral history suggests that the population of New Georgia in 1800 may have been twice that of today, with settlements of up to 7000 people. Thus, this apparently pristine island rainforest may be a consequence of regeneration over the past 150 years, following population decline and migration to coastal regions (11). The enrichment of these present-day island rainforests in secondary tree species adds further support to this possibility.

Beyond pure interest in the antiquity of tropical rainforests and how humans may have shaped them, these studies have important implications for conservation biology. First, they indicate that it is no longer acceptable to suggest that land loss caused by previous human activities was too small to have had a major impact and therefore cannot be compared to present-day losses. This point is reinforced by current estimates suggesting that the main agent of deforestation in the tropical forests, accounting for up to two-thirds of the annual loss, is slash-and-burn farming (12). Thus, although the rate and extent of forest clearance might be much greater today, in many cases, the process is comparable to prehistoric losses. Second, in most examples of prehistoric disturbance there was subsequent forest regeneration. These tropical ecosystems are not as fragile as often portrayed and in fact are quite resilient. Left for long enough, forests will almost certainly regenerate. Third, a long-term paleoecological record enables realistic estimates to be made about the time it takes for a forest to regenerate after abandonment. It also provides a quantitative assessment of forest composition before and after disturbance, providing detailed information on what is missing from, or added to, the forest when it regenerates. Such information is critical to the management and long-term conservation of newly cleared regions and should be routinely incorporated into restoration planning. These approaches have been successfully applied to conservation of mid-latitude forests (13) but have yet to be applied systematically to some of the most diverse but threatened forests on Earth.

References and Notes

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
View Abstract

Navigate This Article