Tải bản đầy đủ - 0trang
V. Weathering Release of Nutrient Elements from Soil Minerals
M. L. JACKSON AND Q. DONALD SHERMAN
plants and systems of soil management which would exploit with maximum efficiency the nutrient elements released from soil minerals by
chemical weathering. Improvement in rooting vigor and selection of
varieties which are “tolerant to” (i.e., yield well in spite of) soil acidity
or low nutrient levels, for example, are ways of exploiting this source of
nutrients. Crop rotation, patch agriculture (in the tropics), and paddy
culture are other examples. Soil treatments have been devised to enhance
the availability of nutrient ions of native soil minerals, In the latter
category particularly has been the treatment of the soil with ground
limestone to release some of the phosphorus tied u p with the trivalent
1. Qemeral Levd of Soil Fertility I s Related t o Weathering Stnge
In both the temperate and tropical zones, the general level of soil
fertility is determined by the weathering stage of the minerals present.
This idea is often expressed in terms such as “inherently fertile soils,”
“old infertile soils,” “highly leached soils,” and even “senile” and
“dead soils.” Soils containing minerals of the early weathering stages,
such a s those which have recently undergone glaciation in the temperate
zone, soils that have recently received ash fall in the tropical zone and
soils that continually receive alluvium containing minerals of intermediate or early weathering stages, are generally recognized as being fertile
a. Influence of Bineral Content in the Temperate Zon-e. Fertility
of soils in the temperate zone is dependent not only on relief, organic
content, and texture but also on the kinds of minerals in the soil. Baren
(1934, 1935) concluded that the nutrient supplies of soils of the Netherlands were correlated with the content of feldspars and biotite. A study
by Hawkins and Graham (1951) of ten widely different Missouri soils
showed that the percentages of potassium feldspars and plagioclase feldspars occurring in the silt was in direct proportion to the fertility levels
and in inverse proportion to the stage of weathering. I n Wisconsin and
Indiana, soils formed on the older tills are much less fertile than those
formed on the latest glacial till. For example, the potassium supplying
power was only one-third as high for the Almena soil (early Wisconsin
till) as for Miami soil (middle Wisconsin till), according to work of
Evans and Attoe (1948).
b. In-fluenoe of Mineral Content in the Tropical Zone. I n tropical
soils, the weathering of the minerals progresses a t a rapid rate. The
bases are released in rather large quantities in the early stages of weathering. Baren (1941) concluded that the fertility of tropical soils of
Sumatra was correlated with their content of primaxy minerals. Mohr
CHEMICAL WEATHERING OF MINERALS IN SOILS
(1944) has divided weathering into five stages : fresh, juvenile, virile,
senile, and laterite stages. Plant growth increases rapidly in the juvenile stage and reaches its maximum in the early part of the virile stage,
because a t this point the nutrient elements are being released to the soil
solution a t a maximum rate and the level of base saturation of the soil
is a t a maximum. The soil according to Mohr is classified as a brown
earth. These soils have a high content of 2 : l layer silicate minerals.
Plant growth begins to decrease at the end of the virile stage and decreases rapidly as the senile stage develops. The end of the virile stage
is approximately a t the point one would expect to find the peak of kaolinization. From that point on, the kaolin is succeeded by the hydrated free
oxides. Plant growth all but ceases with the development of the laterite
The natural productivity of the soils in the early stages of weathering
has been observed by other workers. Hardy (1939) states that volcanic
ash soils develop as rapidly as or more rapidly than they erode. Their
productivity is high. Edelman and Beers (1939) have reported that
infertile soils can be rejuvenated by application of volcanic ash. Edelman (1946) states that the young soils on both volcanic ash and lava are
very fertile. Edelman and Beukering (1948) state that in the tropics
the young soils developed on volcanic ash are rich. Chevalier (1949)
describes a similar relationship in Africa to that described by Mohr in
I n general the most productive soils in the tropical regions are the
brown soils and the less weathered of the tropical red loams. I n these
soils, the chemical weathering stage passes from soils having a mixture of
montmorin and kaolin minerals to the peak of kaolinization. It is a t this
point in soil development where one obtains the best combination of high
base saturation and physical properties. It is on this type of soil that
the great agricultural development has occurred in the Hawaiian Islands.
This would be comparable to Mohr’s (1944) point of maximum growth.
As leaching progresses, with further mineral weathering, the supply
of available nutrient elements decreases to a point a t which plant growth
becomes limited. Aubreville (1947), Chevalier (1948, 1949), Pendleton
(1941), Scaetta (1938), and Edelman (1946) describe the decrease i n
vegetation which develops during the senile stage of soil development
and lack of vegetation on the laterite crust. In almost every case the
loss of vegetation precedes the formation of the crust, and thus the loss
of vegetation must be partly due to the low level of plant nutrients. It
is true also that the physical properties of the crust are not conducive
to plant growth.
M. L. JACKSON AND Q. DONALD SHERMAN
2. Weathering Release of Nutrient Ions from #oil Minerals
The nutrient ions to be considered in connection with weathering
release from soil minerals include potassium, calcium, and magnesium
among the common cations (and associated nickel and chromium among
the toxic elements) ; phosphorus, nitrogen, and sulfur among the common anions ; iron, manganese, copper, zinc, boron, and molybdenum
among the minor elements.
a. Weathering Release of Potassium. The chemistry of soil potassium
has been reviewed through 1951 previously in this publication (Reitemeier, 1951) and elsewhere (Reitemeier et al., 1951). The water collected from lysimeters (Demolon and Bastisse, 1946) filled with sieved
granite showed more potash released per year than those filled with soils.
Since fresh rock has a low base exchange capacity, little of the released
potash could be held in exchangeable form as it is held by soil.
Rouse and Bertramson (1950) showed a correlation of the 10-A.
diffraction intensity (mica content) of soil clays with the potassium supplying power of several Indiana soils, and Phillippe and White (1952)
showed a correlation of microcline content of the silt with the acid-soluble
potassium level in twelve Indiana soils. Expansion of mica interlayer
spaces with weathering has been discussed in Section 11,l c in connection
with the weathering sequence, whereby vermiculite and montmorin appeared to weather from mica. Barshad (1948, 1950) showed in the
laboratory that magnesium, calcium, and sodium would open up the
interlayers of mica>to the 14-A. vermiculite spacing by slow exchange,
and thus demonstrated a likely mechanism of release of mica potassium
to crops in available form. White (1951) showed that treatment of an
illite with magnesium salt solutions gave rise to a 17.8-A. spacing (in
the presence of a. lessened 10-A. diffraction intensity), thereby further
strengthening evidence of montmorin formation during potassium release
from some kinds of mica.
P r a t t (1952) showed a high release of potassium from the coarse clay
fraction of thirteen Iowa soils ; the mica content is known to be high in
the coarse clay fraction of soils in that locality. Kolterman and Truog
(1953) found that large amounts of potassium could be released to exchangeable form by the heating of soils and micaceous clays to 300" C.
while saturated with ammonium. Graham and Turley (1948) showed
weathering release of nonexchangeable potassium to forms available to
Mare1 (1947) found that andesine occurring in the clayey andesitic
soils of the coastal regions of Jav a could deliver sufficient potassium to
sugar cane, which removes about 200-250 kg. of KzO per hectare in
CHEMICAL WEATHERING OF MINERALS IN SOILS
about 130 tons of stalks. I n the tea district of Korintji (west coast of
Sumatra) situated on soils of andesitic basaltic origin, potassium deficiency disease appeared in the plantations. The reason is that soils
developed from basic rock are so poor in potassium that they can not
deliver even 40 kg. KzO per hectare, which is needed to obtain an annual yield of about 1,400 kg. of dry leaves per hectare. These soils,
however, contain plenty of calcium, magnesium, and phosphoric acid.
Soils of acid rock origin, such as rhyolitic soils, however, are rich in potassium but poor in calcium, magnesium, and phosphoric acid.
Michio Ota (Tokyo University) studied the agricultural values of
certain minerals containing potassium, such as biotite, orthoclase, green
tuff, and jarosite (unpublished). Among these materials, he found that
jarosite shows the most agricultural value in a paddy soil in the summer.
The jarosite used is found in association with a limonite deposit of the
Suwa Mine, Nagano Prefecture, formed in a hot spring, and its potassium is not soluble in water. Ota confirmed that hydrogen sulfide, which
is commonly produced by the reducing environment in a paddy soil, acted
upon jarosite, combined with the iron of this mineral, accelerating its
decomposition and releasing its potassium.
b. Weathering Release of Calcium, Magnesium, Chromium, and
Nickel. Calcium and magnesium release from limestone and dolomite
subsoils is a matter of common knowledge and frequently prevents deficiencies of these elements in soils which are otherwise low in exchangeable calcium and magnesium. Graham (1941a) showed that hornblende
and augite released calcium at a rate thought to be sufficient for soybeans.
Marel (1947) found that Feathering of amphibole, hyperthene, and biotite supplied sufficient calcium and magnesium for crop growth in
Sumatra. Marel (1950b) observed that weathering of minerals in temperate climate in the Netherlands is much slower than he had noted in
the tropics. He found, however, that in the climate of the Netherlands,
dolomite could supply enough magnesium and glauconite, enough potassium and magnesium to oats (cultivated in Mitscherlich pots) to obtain
optimum harvests. Biotite could not supply sufficient potassium and
magnesium in this climate to obtain optimal growth of the oats in the
The release of calcium by weathering of anorthite and other plagioclases was shown by Graham (1941a, 1941b, 1942). The weathering rate
was shown to be hastened by the presence of hydrogen-saturated clays,
and was significant to soybean nutrition. The p H of hydrogen clay rose
in the presence of anorthite from 3.3 to 5.7 in 107 days. The amounts of
nutrient release had also previously been demonstrated to be significant
to plant growth (Albrecht e t al., 1938; Graham, 1939).
M. L. JACKSON AND 0. DONALD SHERMAN
Release of magnesium from soils high in the mineral chlorite gives
rise to excess of magnesium as compared to calcium, the areas sometimes
being known as the “serpentine barrens.” The literature on this subject
is reviewed by Robinson ef al. (1935) ; their analysis of soils and plants
indicate toxicity of chromium and nickel in many infertile serpentine
soils. Molybdenum deficiency is also associated with the serpentine barren soils (Walker, 1948; R. P. Stout, lecture a t the University of Wisconsin, 1952).
Chromium is often found in very appreciable quantities in the
tropical ferruginous soils. Simpson (1916) found the content of chromium oxide of ten western Australian laterites to range from nil to 5.3
per cent. Soils having a high content of chromium have been reported
in South Africa by Merwe and Anderssen (1937) and in Samoa by Birre11 et d . (1939). I n the former, the soluble chromium is so high in the
soil as to be toxic to plants.
Chang and Sherman (1953) have found evidence to support the accumulation of appreciable amounts of nickel in the soils of the Low
Humic Latosols of the Hawaiian Islands. The nickel has been released
in the weathering of the olivine-rich basalts. The available nickel content
of these soils is high and approaches levels which may be toxic to plants.
c. Weatheriwg Release of Phosphorus. The native phosphorus of
soils, aside from that in the organic matter, which may be half to
three-fourths of the total, occurs in soils largely a s the mineral apatite,
(C a l o( P 04) a ( F,0 H)z,and as a family of iron and aluminum hydroxy
phosphates. As weathering of the mineral phosphates proceeds, small
increments of the phosphorus go over into solution and adsorbed forms
from which crops take it. Moderately rapid release of native mineral
phosphate appears to take place in soils of very slightly acid reaction
(Truog, 1916). The calcium phosphate of alkaline soils is only slowly
available, except as the soil is acidified by the application of sulfur or
other acid-forming material. Mare1 (1947) found that volcanic glass,
hypersthene, and anorthite contained sufficient apatite inclusions and
weathered fast enough to supply enough phosphorus for crops in
Phosphorus in hydroxy salts of ahminum and iron appears to be
very slowly released to crops if the soil is highly acid ; however, when a n
acid soil is amended by the application of finely ground limestone or the
acidity decreased in other ways, phosphorus is often released from minerals a t a significant rate for crop growth. Snider (1934) reported that
the addition of limestone to certain Illinois soils increased the solubility
of the native phosphorus to such an extent that no additional yield responses were obtained from phosphate applications. Salter and Barnes
CHEMICAL WEATHERING OF MINERALS IN SOILS
(1935), working in Ohio, reported constant wheat yields and increasing
corn yields throughout thirty-two years without phosphate applications
on plots which received adequate limestone, potash, and nitrogen fertilization in addition to manure. Without limestone addition, the yields
steadily declined with the same treatments. Cook (1935) found that
more phosphorus was extractable in dilute acid after several acid soils
were limed than before. Dean (1938) reported that the amount of alkalisoluble inorganic phosphorus in some soils decreased significantly with
increase in pII, but that the amount of weak acid-soluble (mineral) phosphorus increased with pH. Unpublished studies a t the University of
Wisconsin by 0. J. Attoe and associates also indicate a marked increase
in the availability of soil phosphorus as a result of liming acid soils.
d. Weathering Release of Nitrogen and Sulfur. Although the release
of nitrogen is for the most part associated with organic matter of soils
and is not concerned with mineral weathering, two aspects deserve mention. Nitrogen in fixed form occurs in igneous rocks in some abundance
(Ingols and Navarre, 1952). Leaching of it in mountain springs yields
an appreciable amount of fixed nitrogen. Although NH4Cl occurs in
volcanic materials, this is not the only source in igneous rocks, because
fixed nitrogen is yielded by weathering granitic rocks long after soluble
nitrogen would have been leached. Occurrence of ammonium ion in
micas and vermiculites (Bower, 1951) could well account for the slow
release of nitrogen salts from both basic and acidic rocks, as well as for
the fixation of soil ammonium in slowly available mineral form.
Besides its occurrence as soluble salts in soils, sulfate occurs in soils
as the moderately soluble mineral, gypsum ( CaS04* 2 H z 0 ) , and the
slowly soluble mineral, barite (BaS04). The latter mineral has been
associated with infertility of soils by Robinson et al. (1950). Also it
has been shown to be much more soluble in the presence of layer silicate
colloids than in water (Bradfield, 1932). This is explained by the fact
that the barium ion activity is decreased by the colloid to values lower
than would be expected in a solution of the ions in contact with barium
sulfate. This type of action by colloidal clay on barium sulfate solubility is closely akin to the action of hydrogen clay in weathering minerals,
as shown by Graham (1941a, 1941b, 1942).
e . Wea#herilzg Release of Minor Elements. The weathering release
of minor elements will be considered in relation to copper, zinc, iron,
manganese, boron, and molybdenum. The possible relationships of soil
minerals to the trace element status of soils have been reported. Thomas
(1940) suggested that copper and cobalt are ordinarily associated with
the more basic magmatic constituents, and that they are relatively rare
in the feldspathic and quartzose minerals of the more acid igneous rocks.
M. L. JACKSON AND
Association of copper supply with soil mineral weathering is suggested
by the frequent association of copper deficiency with organic soils
Wager and Mitchell (1943, 1950) indicated that the various trace
elements are associated with several stages i n differentiation and consolidation of the original rock material. Chromium and nickel are concentrated in the first rocks to solidify; vanadium in the early middle
stages; copper and lithium in the later middle stages; and molybdenum,
zirconium, thorium, lanthanum, and rubdium are concentrated in the last
rock to solidity, i.e., granitic types.
Graham (1953) studied the relationship between soil mineral weathering and the trace element status of some Australian soils. H e reported
that the stage of weathering, as established by analysis for quartz, feldspar, and heavy minerals, was advanced for soils of known and suspected
trace element deficiency, and was moderate to low in soils nondeficient in
trace elements. High amounts of gibbsite were found in the 2- to 20-y
fraction of soils from areas demonstrated to be copper-deficient. Gibbsite was not found in any of the soils nondeficient in copper. Carroll
(1944) noted a good correlation of the ferromagnesian mineral content
of the sand fraction of some Australian soils with high to very high copper contents of subterranean clover grown on these soils. The occurrence
of gibbsite in soils of demonstrated copper deficiency suggests, according
to Graham (personal communication), the importance of minerals which
could render some heavy metal plant nutrients unavailable, a subject
which he believes might well be investigated further.
The weathering rates of several copper minerals (chalcopyrite,
CuFeSz ; bornite, Cu6FeS4; copper glance, Cu2S; malachite, CuC08
C U ( O H ) ~were
investigated (Steenbjerg, 1943, 1951) by means of crop
yield curves and by measurement of the copper absorbed by barley plants
grown in pots in a soil very deficient in copper. The relative number
of copper atoms on the surfaces of the minerals added to the soil were
calculated. The size of applications was adjusted to give the same number of surface copper atoms in the different copper minerals. In this
way it is possible to observe how plants react to copper added in copper
minerals with different crystal structures.
It was shown that there exist two main types of weathering of sparingly soluble fertilizers (minerals) added to, or occurring in, the soil.
The two types are termed the “copper type” and the “potassium type.’’
Owing to small absorption and absence of leaching, the products of the
copper type of weathering accumulate in the soil, and the physical and
chemical properties of the newly formed compound or compounds rather
than those of the original mineral become the dominant factor in nutrient
CHEMICAL WEATHERING OF MINERALS I N SOILS
availability to plants. On the other hand, with the potassium type of
weathering, particularly of E minerals or slightly soluble N compounds,
the products are rapidly removed from the mineral added owing to the
high degree of absorption and leaching ; under these conditions, it is the
crystal structure and weathering rate of the original mineral that detexmine nutrient absorption and dry matter production.
Zinc is thought to occur in soil minerals mainly as an isomorphously
substituted cation in an octahedrally co-ordinated position normally
occupied by magnesium (Elgabaly and Jenny, 1943 j Elgabaly, 1950).
An association of zinc deficiency with the mineral content of soils is
suggested by the occurrence of zinc deficiency on organic soils (Ellis,
Iron occurs in soils in the minerals hematite, ilmenite, magnetite, and
goethite; in x-amorphous hydrous oxide colloids and coatings; and in
isomorphously substituted positions in micas and other 2 : 1 layer silicates. Availability of iron to crops rests largely on its being reduced to
the more soluble ferrous form, just as its migration during weathering
depends to a considerable extent on its being reduced, Eliman (1938)
showed that plants had a means of reducing iron in the environment of
the root hairs. Chapman (1939) demonstrated the release of iron from
finely ground magnetite through its uptake by citrus seedlings.
I n the Hawaiian Islands, iron deficiency occurs on some soils even
though they contain large quantities of free iron oxides. According to
Johnson (1917, 1924), the high content of manganese dioxide in the
soils prevents the assimilation of iron by the plants. It is suggested that
manganese of the soil oxidized the iron to the ferric form which was unavailable to plants. Sherman and Fujimoto (1946, 1947) have evidence
which attributes the iron deficiency of the plants to an unfavorable manganese : iron ratio in the plant. They point out that both manganese and
iron exist in an oxidized state in the soil and that conditions other than
the content of manganese in the soil control the oxidation-reduction environment of the soil.
Manganese occurs in the soil as hydrous manganese oxides, pyrolusite, manganite, and braunite. The solubility of the various manganese
minerals in the soil is very low, especially that of the hydrous manganese
oxides. According to Sherman and Harmer (1943) , the plant can utilize
only the manganous ion, and thus available manganese will include the
water-soluble manganese, the manganese adsorbed to the soil colloids,
and the manganic manganese which can be reduced a t the root surface.
Fujimoto and Sherman (1948) have found that the available manganous
ion will increase in the soil with an increase in conditions favoring reduction of manganic ion or dehydration of the colloidal hydrated manganese
M. L. JAUKSON A N D Q. DONALD SHFJtMAN
oxides. Leeper (1935) has classified the manganese in soils into the following forms : (a) manganous manganese ; (b) the colloidal hydrated
manganese dioxide; and (c) inert manganese dioxide. H e proposed a
hypothesis in which the soil manganese would exist in a n equilibrium
yhich can be expressed by the following equation:
Manganous manganese # hydrated colloidal MnO2 # inert MnOZ.
Leeper (1947) has modified this equation to include Mn203, MnsOs, and
MnO2 in the term “manganic oxides.” Sherman and Harmer (1943)
have shown that there is an equilibrium between manganous and manganic manganese in the soil and that the direction of the equilibrium is
determined by soil reaction, the oxidation-reduction conditions of the
soil, and soil moisture conditions. Likewise, Fujimoto and Sherman
(1948) have concluded from their work that two processes influence the
availability of manganese in the soil other than soil reaction and that
they are, first, the oxidation-reduction conditions and, secondly, the conditions for hydration and dehydration of the manganese oxides. I n very
acid soils, hydrated manganese oxides will not form, If the soil solution
contains any appreciable concentration of hydroxyl ions, the hydrated
oxides will form readily in the manner described by Bertrand and others
whose work is quoted in Mellor (1932). Thus acid soils will have a high
content of available manganese, and alkaline soils will have a low content of available manganese. Sherman and Harmer (1943) and
Fujimoto and Sherman (1948) have shown that if the hydrated
colloidal manganese oxides are dehydrated, they become soluble. It is
proposed that the dehydrated manganese oxides (MnO).. (Mn03)y(HzO), break down to its component parts. These workers feel that the
release of manganese in the soil is due to physical and chemical conditions. Leeper (1947) and Dion and Mann (1946) have presented data
to show that biological factors may be responsible for the reduction of
manganic manganese in the soil. Fujimoto and Sherman (1948) and
Dion and Mann (1946) have developed manganese cycles in soils which
are very similar, except that the former based their cycle on the physical
and chemical conditions of the soil, whereas the latter based their cycle
on biological conditions.
In certain soils there is a definite accumulation of pyrolusite in concretionary form. According to Sherman et al. (1949b), concretions of
pyrolusite will form in predominantly dry soils which are appreciably
wet during some season of the year. These workers have found that the
manganese is reduced during the wet season and is precipitated during
the dry season. The concretionary deposition occurs around small nuclei,
pores or small holes in soil, around aggregates, and on the surface of
CKEMICAL WEATEIERING OF MINERALS I N SOILS
roots. Living plants with large pyrolusite concretions around the root
were found by these workers in the Hawaiian Islands.
The chemistry of soil boron has been reviewed previously in this publication (Berger, 1949) ; the consideration of its release by weathering
will therefore be brief. Of the total boron in soils, approximately half
was attributed to tourmaline and allied resistant borosilicates (Whetstone et al., 1942). The other half is fixed in mineral and organic forms
-forms which are more closely related to the boron supply important
to crops. Whetstone et al. (1942) estimated the tourmaline in soils on
the basis of the soil boron insohble in boiling 85 per cent phosphoric
acid. The 200 soils analyzed averaged 30 p.p.m. of total boron, of which
17 p.p.m. of boron was insoluble in the acid treatment and attributed to
tourmaline. Weathering release of boron from tourmaline and like minerals of this degree of insolubility would be expected to be slow.
Soils of known molybdenum deficiency and suspected molybdenum
deficiencies were characterized (Graham, 1953) as being highly weathered. They contained a very high percentage of quartz and extremely
low amounts of feldspars. A very low amount of ferromagnesium and
ferrocalcium minerals was found in some of the samples. However,
molybdenum-deficient soils in general were characterized by a very small
amount of weatherable minerals in the heavy fraction.
3. S y s t e m s of Utilization of Soil Mineral Weathem‘ng Sources of
The agricultural development in the tropical as in the temperate
regions has been along both primitive and modern lines. There is a tremendous contrast between the intensive agricultural crop production
found in the Hawaiian Islands and the Dutch East Indies by the addition
of mineral fertilizers, and the “patch agriculture” practiced so extensively in the tropical regions. There are two entirely different philosophies in soil management behind these very different systems of
agricultural production. Although most people think of the tropical
agriculture as being the ultimate of self-sufficiency, it is actually fraught
with many problems of crop nutrition. One must realize that soils in
the tropics range from youthful to senile in their stage of development.
The youthful soils such as the soils of the “red and black complex” and
tropical red loams are productive because of their high base status and
the presence of the montmorin and kaolin minerals.
I n the old soils of the humid tropics, the condition is very different.
Under these conditions the soils are progressively losing their bases.
The progress of this leaching process is retarded by the native vegetation in that i t is constantly returning the bases from the subsoil to the
M. L. JACKSON AND GI. DONALD SHERMAN
surface by the fall of leaves to the surface. As long as the forest can
maintain an adequate base circulation from the subsoil, through the tree,
to the leaves and finally to surface soil in the leaf fall, the soil remains
productive. When the forest is removed the whole cycle is interrupted
with the result that marked changes occur in the physical and chemical
properties of the soil. Thus the activities of man, according to Setzer
(1949), should be included in the soil-forming factom of tropical soils.
With the exposure of the soil, leaching of the remaining bases is accelerated; the decomposition of the organic matter is rapid; and the physical condition of the soil deteriorates.
a. Patch Agriculture. One of the most common forms of agricultural
development in the tropical areas is the so-called patch agriculture. The
people learned from experience that by removing the forest by either
clean cutting or burning they could produce several good crops. As
crop yields deteriorated rapidly and became unprofitable, a new area was
cleared and the old area was allowed to reforest. The basic concept was
that the clear cutting or burning of the forest would cause a release of
nutrients in the soil owing to the rapid decomposition of the humus.
Soil exhaustion set in after this supply of readily available nutrients had
been utilized by the plants o r had been leached from the soil. Reforestation would rejuvenate the soil so that the process could in time be repeated. Ra,witscher (1946), studying the exhaustion of tropical soils,
concluded that deforestation leads to increased leaching and removal of
essential bases, thus leading to the reduced crop yields.
The depletion of soil nutrients was not the only cause for lower crop
production. The exposure of the soil caused the soil structure to deteriorate by the development of the Laterite crust. Humbert (1949) states
that the main causes of soil exhaustion under this system of agriculture
are as follows: (a) clear cutting in the forest; (b) fires; and (c) shifting
cultivation practice in the forest. He states that it takes fifty years for
the laterite crust t o form after the removal of the forest. Chevalier
(1949) has observed tha,t with the removal of the forest, the laterite
clays quickly become senile with the development of iron concretions and
iron-oxide pans below the surface. The crust eventually appears a t the
surface, and the area becomes barren and unfit for the growth of plants.
It takes trees forty years to restore productivity of the baxren areas
(bovals). Chevalier has found that laterite crust forms rapidly in New
Guinea. Aubert (1950) considers crust formation as being a slower
process, and according to his estimates it takes sixty years for the crust
to form. Aubert (1949)) Aubreville (1947,1948), and Chevalier (1948)
attribute the loss of productivity to the formation of the laterite crust
due to its poor physical condition. Sherman et al. (1953) have found