Environmental Engineering Reference
In-Depth Information
for the global forest area even for the twentieth century. All of the estimates that
try to reconstruct even longer periods of the Earth's changing forest cover have even
greater error margins: they must be seen merely as useful indicators of likely trends,
not as markers suitable for solid quantitative conclusions.
Differences among the chosen dei nitions of forest account for most of the uncer-
tainties in reconstructing the preagricultural extent of forest biomes and calculating
the degree of their subsequent reduction. Matthews (1983) set the global area of
preagricultural closed forests and woodlands at 6.15 Gha and estimated that by
1980 it was cut by less than 1 Gha, to 5.2 Gha, a reduction of only about 15%. In
contrast, Richards (1990) concluded that 1.39 Gha of forests were cleared in less
than 300 years between 1700 and 1995, with at least 950 Mha (nearly 70%) in the
tropics and less than a third in temperate ecosystems. Williams (1990) estimated the
pre-1650 global forest clearings at just 99-152 Mha, followed by 60-75 Mha
between 1650 and 1749, 159-168 Mha between 1750 and 1849, and 418 Mha
between 1850 and 1978, for a cumulative total of 740-810 Mha. Ramankutty and
Foley (1999) assumed that the potential forest and woodland vegetation covered
55.27 Mkm
2
, that it was reduced by less than 5% by 1700 (to 52.77 Mkm
2
), and
that in 1992 it covered 43.92 Mkm
2
(4.39 Gha), roughly a 20% reduction of the
potential cover.
Discrepancies in estimates of forest areas do not end with historical reconstruc-
tions: contemporary i gures also have a surprisingly large range of uncertainty.
Whittaker and Likens (1975) put the forested area at 50 Mkm
2
, while Matthews
(1983) settled on 33.51 Mkm
2
of forests and 15.23 Mkm
2
of woodland, and later
chose a more restrictive dei nitions of forest, limiting it to areas of closed or partially
closed canopies, resulting in a total of only 29-34 Mkm
2
(Matthews et al. 2000).
Olson, Watts, and Allison (1983) listed 30.8 Mkm
2
of tree formations (woods) and
27.38 Mkm
2
of interrupted woods; Ajtay et al. (1979) had only 33.3 Mkm
2
in
tropical, temperate, and boreal forests and another 2.5 Mkm
2
in Mediterranean
shrublands, while Saugier, Roy, and Mooney (2001) put the totals in the same two
categories at 41.6 and 2.8 Mkm
2
. None of these totals agrees with the FAO's recent
global assessments.
Estimates of historical rates of deforestation are even more questionable. There
are no reliable global estimates of deforestation during antiquity and the Middle
Ages. This absence holds even for China, a country with an exceptionally long
record of local and county gazetteers that were used to record all kinds of events
and changes. Such fragmentary and discontinuous information makes it impossible
to reconstruct the advancing deforestation on a nationwide scale. Wang (1998 esti-
