Geography Reference
In-Depth Information
Box 14.2
No acidification in Asia's mountains
Why is it that Asia's tropical forest catchments suffer acid
rain, but not stream water acidification or acid-induced
deforestation like that experienced in the West? It is well
established that trees are effective scavengers of acidic
atmospheric compounds and that this process increases
acid deposition in forest catchments to levels several times
those experienced in open fields (Neal
et al
. 1992).
However, despite receiving inputs similar to basins in
industrial Europe and in defiance of years of careful
monitoring, Japan's forest basins do not show signs of
acidification. Review of the international literature suggests
that they are not alone. Similar findings come from other
mountain basins in Asia (Ohte
et al
. 1998). Indeed, there
seems to be a geographical regionalisation of the way
catchments respond to acid deposition. In Asian
experimental basins, rainwater of pH 4 to 6 is converted to
stream water of pH 6 to 7. In North American and
European (except Mediterranean) basins, rainwater of pH
4 to 5 is converted to stream water of pH 4 to 6
(ibid.).
Investigations at Kirya watershed, Japan, suggest that
its hydrochemical processes are controlled by two factors:
a biological factor and a geochemical factor
(cf.
Finley and
Drever 1997). Biological reactions in the upper soil
increase the pH from 5.5 to 6.0 during the unsaturated
vertical infiltration of throughfall and temporary saturated
lateral infiltration from the hill slope. Two mechanisms are
active: cation (Ca
2+
and Mg
2+
) exchange with organic acids
near the soil surface, and proton consumption by
weathering reactions, using acid supplied by the CO
2
dissolution-dissociation reaction. Thus the acidity, gained
from throughfall and from the production of organic acids
in the soil, is mitigated by exchanges in the soil and during
infiltration to groundwater.
The balance between these processes is affected by
forest succession (Asano
et al
. 1998). Study of three
steeply sloping (average 20-34°) basins on granite
bedrock confirms that the major hydrochemical
controls are differences in the interactions between
biological and geochemical factors. Biological NO
3-
is
the major source of H
+
under the mature forest,
geochemical SO
4
2-
under the younger forest, while pH
changes little during infiltration in the deforested
control basin. There was an imbalance between cation
supply and demand in the younger forest basin, where
the trees are active in the acceleration of geochemical
weathering.
However, neither tree growth nor geological factors
can explain the differences between these basins and
those of Europe. The Japanese researchers suggest
that the cause lies in the relative immaturity of the soils
in the Japanese and other Asian mountain headwaters.
Ohte
et al
. (1998) point out that as forest soils develop,
the cation-leached layer becomes thicker and the
unleached layer with high buffering capacity becomes
thinner. Typically, Europe's podsolic soils generate acid
from deep surface organic accumulations, and
infiltration proceeds through a soil horizon that is
entirely leached. Asian forest catchments on steep
hillsides preserve immature soil profiles through soil
refreshment by erosion and deposition. In Asia's steep
lands, erosion and landsliding often bury or eliminate
litter accumulations. In addition, in many young forest
soils, the infiltration layer includes a rapidly developing
C-horizon of weathered but not yet much leached
bedrock. The final pH of the infiltrated water depends
on the relative effectiveness of the acid-producing litter
layer and the unleached C-horizon. The differences in
buffering capacity between the Western and Asian case
studies may thus be related to local slope steepness,
slope instability and the age of the forest soil.
almost nothing, and runoff climbing to 190 m
3
ha
-1
month
-1
from 2 m
3
ha
-1
month
-1
on newly
constructed log skid roads. However, soil and runoff
losses from new agricultural lands may also be high.
In Sri Lanka, where agricultural extension reduced
forest cover from 70 per cent to 20 per cent
between 1900 and 1995, there was an increase in
the rainfall/runoff ratio estimated as 0.7-1.4 per
cent per year. In areas cultivated to tobacco and
upland annual crops, soil loss became 25-70 times
that in forest (Haigh
et al
. 1998). Deforestation is
generally associated with dramatic increases in soil
loss and sediment yield (Derose et al. 1993).
A review of about eighty studies of surface
erosion in natural forest and tree-crop systems
finds median soil loss in natural forests to run at
about 0.3 (range 0.03-6.2) Mg ha
-1
yr
-1
(Wiersum
1984). Plantations and tree crops with ground
cover/mulch suffer losses of about 0.6-0.8 (range
0.02-6.2) Mg ha
-1
yr
-1
. A small number of studies
of the cropping phase under long-rotation forest
fallow suggest soil losses of around 2.8 (range 0.4-
70) Mg ha
-1
yr
-1
. In the western Himalaya,
comparison of bedload sediments trapped from
parallel streams draining four, steep 1 km
2
microcatchments found that the sediment loads
from undisturbed forest were five-seven times
smaller than from deforested areas covered by grass
and scrub. The depths of soils on the deforested
areas were significantly smaller, and there were
larger patches of exposed bedrock (Haigh
et al
.
1998; Harding and Ford 1993).
Conversion to arable cropping often accelerates
erosion. On poorly managed agricultural steep
