forest resource assessments,
Question about Long-Term Tropical Forest Decline
Despite widespread concern about tropical forest loss, plenty of local evidence of deforestation and frequent estimates of global deforestation rates, convincing evidence for a long-term decline in overall tropical forest area since 1970 is hard to find.
When I first reviewed this problem in 1980, I predicted, based on available evidence, that tropical moist forests might all disappear in 60 years. (Tropical moist forests are the 'jungles' of the humid tropics, situated near the Equator in Amazonia, Borneo, the Congo Basin, etc.). There have since been frequent estimates of deforestation rates but the long-term trend in forest area has been neglected. My article in the Proceedings of the National Academy of Sciences aims to fill this gap.
The first time series I constructed was for all tropical forest, which includes tropical moist forests and forests in the dry tropics. I used estimates for this statistic published by the UN Food and Agriculture Organization (FAO) in its Forest Resources Assessments (FRAs).
According to FRA 1980, tropical forest covered 1,970 million hectares (mha) in 76 countries in 1980. This was corrected to 1,935 mha in 1982. In FRA 1990, it declined from 1,910 to 1,756 mha between 1980 and 1990 for 90 countries. But in FRA 2000, for the same 90 countries it fell from 1,926 to 1,799 mha between 1990 and 2000. In FRA 2005, it dropped from 1,949 to 1,768 mha between 1990 and 2005.
So while estimates for 1980 have declined over time, those for 1990 have risen. Meanwhile, the trend for 1990 to 2000 almost exactly replicated that for 1980 to 1990.
Inconsistencies between the three trends may result from changes to definitions of FRA statistics, and the large errors generated when combining large numbers of national statistics to give global statistics. National forest surveys focus on commercially valuable tropical moist forests, not forests in the dry tropics, such as savanna woodlands, with a more open canopy. As these 'open forests' account for 40 per cent of all tropical forest, this introduces considerable errors into estimates of total tropical forest area. National surveys are also infrequent, so it is difficult to combine them to give global estimates in the same year (e.g. 2000) without incurring further errors. Producing FRAs is a massive job for FAO's dedicated staff and foresters worldwide.
The correction to the 1990 estimate in FRA 2000 was similar to the amount of forest apparently lost in the 1980s. If this correction is representative of the errors involved in producing global estimates then it raises questions about the reliability of FAO's trends.
My second time series, for tropical moist forests only, removes errors due to open forests and shows no decline. Before 1990 it consists of expert assessments. Afterwards it contains some estimates based on satellite surveys, which reduces combination errors. In the mid-1970s tropical moist forests were thought to occupy 935 mha. In 1983 I estimated that they covered 1,081 million ha in 63 countries in 1980. Two comprehensive satellite surveys by the European Commission's Joint Research Centre lead to estimates of 1,152 mha in 1990 and 1,181 mha in 2000 for these 63 countries.
These findings do not prove that tropical deforestation is not happening, but they do raise questions about our ability to track the overall trend in tropical forest area. Combining many national statistics to give a global estimate involves large errors. The apparent lack of decline in tropical moist forest area may result from more accurate satellite surveys showing more forest than experts thought existed, though using coarse resolution sensors still leads to errors. Deforestation in some places may also be partly offset by natural forest regrowth elsewhere. This is difficult to spot, even on satellite images.
My research raises important questions. For example, if national forest statistics are the only data available should global change scientists use them even if they are not intended to be of scientific quality? If so, should scientists pay greater attention to the errors involved when communicating their findings to fellow scientists and the public? Ultimately, if global change scientists want global data of scientific quality they must establish a World Forest Observatory that can use the full range of satellite data and make ground observations too.
FAO, 1982. Tropical Forest Resources. Forestry Paper No. 30. FAO, Rome.
FAO, 1993. Forest Resources Assessment 1990: Tropical Countries. Forestry Paper No. 112. FAO, Rome.
FAO, 2001. Global Forest Resources Assessment 2000. Forestry Paper No. 140. FAO, Rome.
FAO, 2006. Global Forest Resources Assessment 2005. Forestry Paper No. 147. FAO, Rome.
Grainger A., 1980. The state of the world's tropical forests. The Ecologist 10: 6-54.
Grainger A., 1983. Improving the monitoring of deforestation in the humid tropics. In Sutton S.L., Whitmore T.C. and Chadwick A.C. (eds.), 1983.Tropical Rain Forest-Ecology and Management: 387-395. Blackwell's Scientific Publications, Oxford.
Grainger A., 2008. Difficulties in determining the long-term trend in tropical forest area. Proceedings of the National Academy of Sciences USA 105: 818-823.
Lanly, J.P. (ed.) (1981). Tropical Forest Resources Assessment Project (GEMS) 1980. FAO/UNEP, Rome. [This is surely one of the greatest books on tropical forests ever published.]