Geography Reference
In-Depth Information
lands and sites recently cleared of ground cover,
huge soil losses can be reported. Bruijnzeel (1990)
reports figures from tens to hundreds Mg ha
-1
yr
-1
,
including a maximum of around 500 Mg ha
-1
yr
-1
for fields planted to onions, tilled up-and down-
slope in Java. Soil conservationists consider
recently up- and down-slope filled bare fallow to
be the most erodible condition for an arable field.
The losses from the same land fallowed under
forest may be anything up to three orders of
magnitude smaller, depending on the slope angle,
soil type, climatic regime and scale of the study.
One of the reasons that soil losses increase after
deforestation is that less water soaks into the soil.
This is often due to compaction, which leads to
higher bulk soil densities and reduced infiltration
rates. In the arid Zagros Mountains, Iran,
conversion of a coarse silty calcixerollic xerochrept
soil from oak
(Quercus brontii)
forest to field crops
resulted, after twenty years, in almost a 20 per cent
increase in bulk soil density, a 50 per cent decrease
in organic matter and total nitrogen content, and a
10-15 per cent decrease in soluble ions compared
with soil under undisturbed forest (Hajabbasi
et al
.
1997). In humid Nigeria, studies near Ibadan found
that deforestation and cultivation increased bulk soil
density and penetration resistance but decreased
mean weight diameter of aggregates (Lal 1996).The
infiltration rate declined with deforestation and
time under cultivation. Soon after deforestation,
saturated hydraulic conductivity and equilibrium
infiltration rate in cleared and cultivated land
declined to only 20-30 per cent of that under forest
(ibid.)
.
In Brazil's Rio Grande de Sul, conversion of 90
per cent of the forest allowed soil losses to climb to
a regional level of 7 Mg ha
-1
yr
-1
and was followed
by the emergence of surface flow pathways
(Mendiondo
et al
. 1998; Castro
et al
. 1997). Natural
fallowing resulted in a ten-fold increase in
infiltration rate within five years. Well-managed
agricultural lands had soil losses comparable with
those under natural forest. On steep slopes, the
effect may be more dramatic: cleared sites in the
Western Ghats, India, shed 120 Mg ha
-1
in a single
season. However, erosion rates decline rapidly
subsequently. In the Ghats, erosion from sites
cultivated to pepper were less than 3.5 Mg ha
-1
yr
-1
.
In Rio Grande de Sul, after conversion to no-till
agriculture, soil losses declined to levels too small to
record (Mendiondo
et al
. 1998).
A similar message comes from studies of lands
converted to pasture. De Moraes
et al
. (1996)
examined the consequences of forest conversion
to pasture on soils in the southwestern part of the
Brazilian Amazon basin. After pasture installation,
soil bulk densities were higher in the 0-5 cm soil
layer, with small changes detected in deeper layers.
Soil pH increased, while total soil carbon (0 to 30
cm depth) was 17 to 20 per cent higher in 20-
year-old pastures than in original forest sites.
However, another Rondonian study comparing
soil inorganic N concentrations, net mineralisation
and nitrification rates in forests and pastures (>3
years old) on ultisols/oxisols found that forest soils
had higher extractable NO3-N and total
inorganic N concentrations than pasture soils.
Rates of net N mineralisation and net nitrification
were higher in the forest soils (Neill
et al
. 1997;
Anderson and Spencer 1991: 40). Nitrogen and
nitrification rates remained the same across the six
forest plots tested, suggesting that the controls are
similar. The low mineralisation and nitrification
rates in pasture soils suggest that annual nitrogen
losses from deforested landscapes may be lower
than from the original forest (Neill
et al
. 1997). In
the Barbudal Reserve, Costa Rica, tests have
confirmed that soil C, N and K were lower, while
many base cations and micro-nutrients were
higher in grassland plots than in forest plots
(Johnson andWedin 1997).
The removal of tree biomass by logging
removes on averages 30 and 15 Mg ha
-1
of carbon
from seasonally moist and seasonally dry forests,
respectively. The rotting of roots may add 50 Mg
ha
-1
(Anderson and Spencer 1991). The amount
of C that is lost to the system varies with many
environmental factors as well as with the way the
land is cleared and used subsequently. In soils
under grazed pastures converted from forest 18-
25 years previously were investigated in Costa
Rica. The net loss of C was 2180 g m
2
for a
hapludand and 150 g m
2
for the humitropept soil
(van Dam et al. 1997).
