Tải bản đầy đủ - 0trang
VI. Mobility of Soil Phosphorus
B. MOVEMENTBY MASS FLOW
The amount of movement by mass flow is the product of the concentration of phosphorus in the soil solution and the extent of liquid
flow. This latter may vary considerably, but a discussion of the factors
affecting it are outside the scope of the present paper. Movement by
mass flow may be of importance in bringing soil phosphorus to the
plant root and in causing leaching. Since the concentration of phosphorus
in the soil solution is generally low, the amount of movement will
normally be insignificant; for example, Barber (1962) showed that in
his soils mass flow could account for only a small fraction of the phosphorus taken up by the plants. Similarly, phosphorus is not normally
considered to be lost by leaching, although some loss must occur over
geological time since total soil phosphorus contents are generally lower
than parent materials.
Where the phosphorus concentration in solution is higher, movement by mass flow may be important. In soils of extremely low phosphorus adsorption capacity, for example, Ozanne et al. (1961) demonstrated that phosphorus could leach. Similarly Larsen and Sutton (1966)
showed that considerable phosphorus movement could take place in a
glasshouse soil when the adsorption complex had been satisfied by
heavy phosphorus applications. The phosphorus concentration in the
soil solution may also be raised by the activity of organisms. As Hannapel
et al. (1964b) pointed out, this may account for the large body of
evidence which shows that phosphorus penetration is greater in soils
which have received manure rather than inorganic phosphorus fertilizer.
C. MOVEMENTBY DIFFUSION
Although the study of movement of phosphorus in soil by mass flow
dates back to Way’s classical studies in the middle of the last century,
the study of phosphorus movement by diffusion is of recent origin. It
has advanced only since the advent of 32P,which permits the precise
measurement of movement over short distances.
Diffusion is the process by which matter is transported from one
part of a system to another as a result of the thermal movement of
molecules or ions. This movement is continuous, but where the system
is at equilibrium there is no net transport. However, where differences
in concentration exist, transport will occur, tending to move the system
toward equilibrium. Transport of phosphorus through the soil will
cause chemical changes to occur both in the liquid and solid phases,
which will complicate the measurement of a diffusion coefficient. This
complication can be overcome by using carrier-free 32Pwhich permits
the measurement to be made in the absence of a concentration gradient.
Under these conditions adsorption and desorption will still be occurring,
so that the reactions between the liquid and solid phases must be taken
into account. Since diffusion occurs essentially in the liquid phase and
an individual phosphorus ion spends only a relatively short time in this
phase, the diffusion coefficient of phosphorus in the soil solution will be
different from that in free solution. The important factor is the fraction
of time a given ion spends in the solution phase. A term to account for
this was used by Lewis and Quirk (19sS), who suggested that
where D, is the diffusion coefficient which would have been obtained
if 32Phad not been adsorbed onto the solid phase and D, is the observed
(apparent) diffusion coefficient. In a given soil, the term solution
32P/total 32P and hence D,, will increase with increasing degree of
saturation of the phosphorus adsorption system. Thus Lewis and Quirk
(1965) showed that D, was directly related to the amount of added
phosphate. Comparable terms have also been introduced by other
workers. Olsen et al. (1962) used a capacity factor, related to the
phosphorus adsorption capacity of the soil, which he defined as the
slope of the line relating labile phosphorus to the concentration of
phosphorus in the soil solution.
The inclusion of factors of the type discussed above makes allowance
for only one aspect of phosphorus diffusion in soil. The observed diffusion coefficient will, however, differ from that in free solution in other
respects. An aspect which is of particular importance is the continuity
of the solution phase. This will be influenced by the solid particles
present, by the electrostatic forces in the solution adjacent to these
particles and by the moisture content of the soil. The significance of
these have been discussed by Porter et at. (1960) for chloride ions and
they introduced a “tortuosity factor” in their diffusion theory to allow for
solid impedance as well as factors for volumetric moisture content and
ionic interaction. Olsen et al. (1962) found that the “tortuosity factor”
determined for chloride could equally well be used for phosphorus,
provided that a further correction was made for the adsorption process.
Nye (1966) has pointed out that ions other than phosphate may
influence the rate of phosphorus diffusion. In his development of the
theory of self-diffusion and bulk diffusion in soil, he takes account also
of the small extent to which phosphorus diffusion may occur along the
surfaces of the solid phase.
An evaluation of the fundamental soil properties which are likely
to influence each of the modifying factors would enable predictions of
the rate of phosphorus diffusion to be made. However, the present stage
of development of the diffusion theory relevant to movement of phosphorus in soil is not advanced enough to allow a more complete assessment of these.
Studies on phosphorus diffusion in crumbs, which will yield results
which are relevant to soil in its natural condition, have been conducted
by Gunary ( 1963, 1964) and Gunary et al. (1965). He found that the
rate of diffusion was related to the degree of saturation of the phosphorus adsorption capacity. Thus addition or removal of phosphorus or
treatments which brought about changes in the phosphorus adsorption
capacity of a given system brought about changes in the rate of diffusion.
Complications may arise when the chemical environment within soil
crumbs is considered. In a study of isotopic exchange in a solution in
equilibrium with soil crumbs, Gunary (1963) showed that the 32P:31P
ratio decreased with time despite the absence of diffusion of 32Ptoward
the center of the crumbs. Larsen and Gunary (1965) explained this
observation by a release of 31P to the solution brought about by
anaerobic conditions in the center of the crumbs, which caused a phosphorus concentration gradient between the center and the surface. A
process of this type could be important even at normal soil moisture
levels in that it causes phosphorus to move to a more favorable position
for absorption by plants.
Where soil phosphorus is considered purely as a source of phosphorus
for plants, some simplifications are possible. The precise form in which
the phosphorus exists is then of little significance, and all the emphasis
can be placed on its behavior. Using an extension of Schofields (1955)
analogy, the behavior of phosphorus in soil can be likened to the behavior of water in a well system.
In Fig. 15, the central well is immediately surrounded by highly
porous material, beyond which there is an infinite extent of less porous
material. The level of water in the well will be the same as that in the
high porosity material through which water will flow freely. In contrast
the water in the material of low porosity will flow so slowly that its
level may differ from that in the high porosity material. As can be seen
from the diagram there is a central trough in the bed rock and until
this is full all the water will be confined to the highly porous material.
In the analogy, phosphorus in the soil solution is represented by the free
water in the well, and labile phosphorus by mobile water in the highly
FIG. 15. Well analogy depicting the behavior of phosphorus in soil.
porous material. Nonlabile phosphorus is equivalent to nonmobile
water in the outer zone of low porosity material. The phosphorus
adsorption capacity of the soil will be represented by the amount of
high porosity material, and the width of this zone will consequently
vary when the analogy is applied to different soils.
The parallel between the behavior of water in the well system and
phosphorus in soil can be seen when the two systems are considered
side by side.
When only a small quantity of
phosphorus is present it will be strongly
adsorbed, the concentration in solution
will be below that for the precipitation
of any mineral and all the inorganic
phosphorus will be labile (Larsen,
1964 ) . The concentration of phosphorus
in the soil solution will here be controlled simply by the amount of inorganic phosphorus present and the size
of the adsorption system.
When only a small amount of
water is present it will all be in the
central trough, the level being too low
for seepage into the low-porosity material. In this situation, the level of water
in the highly porous material and in
the well will be controlled by the
amount of water which is present in
the zone of highly porous material and
the width of this zone.
If phosphorus is added, the concentration in solution will rise ,until
an equilibrium level corresponding to
the solubility product of some phosphorus mineral is reached; a crystalline
phase will then precipitate and the
phosphorus within its lattice will no
longer be labile. By further phosphorus
addition, it is possible to raise the concentration above the equilibrium value,
but in time the level will fall until
equilibrium is reattained.
If water is added, the level in the
well and highly porous material will
rise until eventually an equilibrium will
be reached at the lip of the trough.
Seepage of water will then occur into
the low porosity material and nonmobile water will begin to accumulate.
If addition is more rapid than the rate
of seepage, a temporary enriched state
will exist where the level is above the
lip of the trough. Seepage will continue
however, and in time the level in the
well will fall back to its stable equilibrium position.
Conversely, if phosphorus is removed from a soil which has reached the
equilibrium level, the solution and labile
phases will be depleted and nonlabile
phosphorus will be slowly mobilized to
restore the status quo.
Conversely, if water is removed
from a system that has reached equilibrium, the level in the well and high
porosity material will drop and nonmobile water will flow back slowly to
restore the status quo.
This picture of the behavior of phosphorus in soil can be summarized
by the reaction:
in which it will be remembered that the reaction between solution
phosphorus and labile phosphorus is rapid, but that between labile and
nonlabile phosphorus is slow.
The immediate source of phosphorus for plants is the small amount
that is in the soil solution. As this is removed, the equilibrium is disturbed
and phosphorus in the IabiIe fraction will be drawn upon. Nonlabile
phosphorus is not likely to contribute to the supply over a period as
short as one growing season since its rate of release is too slow. The
supply of phosphorus to the plant then depends directly on the concentration in solution and indirectly on soil factors which maintain this.
The factors responsible may be better appreciated by reference to
the well analogy. The concentration of phosphorus in the soil solution
(the level of water in the well) is a function of the amount of labile
phosphorus (amount of mobile water) in relation to the phosphorus
adsorption capacity (quantity of high-porosity material), that is, the
extent to which the sorption capacity is filled which can be expressed
as the percentage saturation. If phosphorus is removed from the solution, it will be replenished from the solid phase labile phosphorus and
the system will readjust to a lower level. If this readjustment occurs
slowly, a temporarily larger drop in the phosphorus concentration in
solution will result. The new level which is eventually attained depends
on the adsorption capacity, since soils with a large adsorption system
will have a greater quantity of labile phosphorus for a given level in
solution. Thus the initial phosphorus level is controlled by the percentage saturation while the buffering of this level is controlled by the
quantity of labile phosphorus.
When appreciable uptake occurs, it will substantially lower the
phosphorus level in the solution immediately adjacent to the roots.
Maintenance of phosphorus supply to the plant will then depend on the
movement of phosphorus to replenish this. Barber (1962) showed that
this movement is primarily a diffusion process, the rate of which is
related to the concentration in solution.
The important factors in phosphorus supply to the plant are therefore the intensity, kinetic, and capacity factors of Wiklander (1951))
and the diffusion factor. The intensity factor is a measure of the concentration of phosphorus in solution; the kinetic factor describes the
rate at which the solution is replenished from the solid phase; the
capacity factor is the quantity of phosphorus capable of replenishing
the solution (the labile phosphorus), and the diffusion factor is the rate
at which the absorption zone is replenished from nearby soil solution.
The supply of phosphorus to plants could be limited by any of these
four factors, and it is therefore of interest to consider their relative
The intensity factor is of direct importance, but account must also
be taken of the extent to which the concentration in solution is buffered. This buffering depends on the quantity of labile phosphorus
present, that is, the capacity factor. For soils with similar adsorption
capacities, the level of phosphorus in solution will be directly related
to the quantity of labile phosphorus so that either the intensity or
capacity factor on its own may be well correlated with plant uptake.
Where a wider range of soils is considered, the adsorption capacities
will vary so that both factors must be taken into account. Thus Gunary
and Sutton (1967) were able to account for 80 to 85 percent of the
variation in phosphorus uptake from a range of soils when intensity
and capacity factors were considered together. This high degree of correlation does not allow for much improvement from further introduction of kinetic and diffusion factors, However, these latter factors have
both been shown to be closely correlated with the concentration of
phosphorus in solution and so were already taken into account by the
intensity factor used by Gunary and Sutton.
That the intensity and capacity factors together can generally fully
describe phosphorus supply in soils can be appreciated from the well
analogy, Here the level of water in the well and the amount of mobile
water with which it is in equilibrium are all that are necessary to
completely describe the short-term water supply for any well system.
For a specific system, the change of status brought about by water
addition or removal can be monitored by following changes in either
one of these parameters.
Similarly, for a particular soil only one parameter need be followed,
and in the subsequent section on the maintenance of phosphorus status,
only changes in the amount of labile phosphorus are considered.
In virgin soils, presumably near to equilibrium, the amount of labile
phosphorus present will be controlled by the solubility product of some
phosphorus mineral. For slightly acid, neutral, and calcareous soils, the
relevant mineral is likely to be hydroxylapatite, so that the concentration of phosphorus in solution will be low.
If phosphorus is removed, the equilibrium level will still tend to be
maintained, since in time, mobilization of nonlabile phosphorus will
occur, as shown for example by Larsen and Sutton (1963) and Vaidyanathan and Talibudeen ( 1965). These latter authors removed phosphorus from soil by means of anion and cation exchange resins. This
treatment brought about an initial decrease in the readily isotopicaIIy
exchangeable phosphorus and they followed the recovery of this fraction
during incubation periods of up to 9 weeks. They were not able to
study in detail the rate at which this recovery occurred, but it appeared
to have been completed within the experimental period. From their
data, the ‘%alf-life”of the process can be estimated to be about 10 days
in one soil rich in isotopically exchangeable phosphorus, and twice as
long in a soil of a lower phosphorus status. This mobilization of pre-
viously nonlabile phosphorus is of agronomic importance in the maintenance of phosphorus levels under extensive agricultural conditions.
For intensive agriculture, the equilibrium level of labile phosphorus
is likely to be far too low for maximum crop growth. The phosphorus
status has to be raised and this may be achieved either by adding more
phosphorus or by reducing the total adsorption capacity. The effect of
reducing the adsorption capacity can be visualized from the analogy,
where reducing the amount of highly porous material without altering
the quantity of mobile water will have the effect of raising the level.
Until more is known of the mechanism of phosphorus adsorption in
soil, progress in reducing the adsorption capacity is bound to be slow.
However, the beneficial effects of organic matter, silicates, and lime on
phosphorus uptake must at least in part be due to blocking or eliminating adsorption sites.
The commonest way of increasing the phosphorus status is by the
addition of phosphorus in manure or fertilizer. Where the initial phosphorus status is very low or the sorption capacity is very high, the
amount of phosphorus required to reach a satisfactory level will be
prohibitive, Under these circumstances it is necessary by fertilizer placement to restrict the amount of soil that the fertilizer actually contacts,
in order for at least part of the growing medium to reach a satisfactory
level. It is well known that the quantity of phosphorus removed by
crops is small in comparison with normal fertilizer additions (recoveries
as low as 10 percent are common). The use of repeated applications
should thus lead to an enriched state.
However, an enriched state is metastable and there will be a gradual
loss of labile phosphorus to a nonlabile form. An exponential rate of loss
of labile phosphorus was suggested by Larsen et al. (1965) and an
example of their results is shown in Fig. 16. In the pH range 5.5 to 7.5
they found half-lives for the rate of fall of labile phosphorus content in
their (mineral) soils to vary from 1 to 6 years, the more rapid loss being
associated with the soils of higher pH. This suggested that the mechanism for the loss of lability could be a slow formation of crystalline
calcium phosphate, presumably hydroxylapatite. The results of Eanes
et al. (1965) suggest that in the formation of pure hydroxylapatite there
is a spontaneous autocatalytic change from the initial amorphous
product to a well crystalline material. It may be that the exponential
curve for loss of labile phosphorus was due to this spontaneous change
occurring at random in the isolated spots where labile amorphous
phosphorus compounds had been formed from the added fertilizer.
The maintenance of an enriched level requires account to be taken
of loss by conversion to nonlabile forms. This loss has been treated
theoretically by Larsen and Probert ( 1968). They considered the
situation where phosphorus that was fully and immediately labile was
added repeatedly in a regular pattern. The loss of labile phosphorus
between -additions would initially be less than that added, and the
phosphorus status would rise. As the status rose the amount lost
o 5001R Pfilacra
FIG. 16. Rate of loss of labile phosphorus after enrichment. Asterisk (left
ordinate): millimoles of P per kilogram of soil. (From Larsen et al., 1965.)
between applications would increase until eventually the loss would
be exactly equal to the amount added. They also considered the situation where the added phosphorus was not immediately labile, for
example, water-insoluble phosphorus. The level attained could still be
predicted provided that the rate of conversion to a labile form (the rate
of dissolution) was known.
For a particular soil the ultimate level attained for any source
depended on the amount of phosphorus added, the interval between
applications, and the rate of loss of labile phosphorus. In the example
shown in Fig. 17, it was assumed that the rate of loss of labile phosphorus and the rate of dissolution of the slow-acting source both obeyed
first-order kinetics. The mean status eventually attained would be virtually independent of source, although the time taken to reach this level
would increase as the rate of dissolution decreased. In the example
quoted it would take many more applications of the slow acting source
than of the fully labile source to reach this situation.
Once a stable situation had been reached the essential difference
between the sources was in the variation which occurred around the
mean status. The water-soluble source resulted in high peaks at the
time of application with subsequent low troughs, whereas the slowrelease source showed smaller oscillations.
Thus when it is required to raise the soil status, it is obvious that the
immediate effect of the water-soluble source is essential. Once a high
status has been reached, its maintenance at a relatively constant level
would require frequent small applications of a water-soluble source,
Water-sobble P source
FIG. 17. Theoretical variation in phosphorus status with time. Equilibrium
situation reached after repeated triennial applications of phosphorus. (Half-life of
immobilization, 2 years; half-life of dissolution of slow-acting source, 1 year.)
whereas less frequent, larger additions of a slow-release source may be
tolerated. However, the contrasting requirements of the various crops
in a rotation require attention, and applications of the water-soluble
source could be phased with advantage so that the peaks coincided
with the most demanding crop.
On such a basis, it should be possible to predict the most suitable
amount, timing, and source of phosphorus for particular agronomic
situations, provided that the relevant soil parameters are known.
The determination of soil phosphorus as a nutrient source for plants
should ideally yield information on the behavior of the phosphorus. It
was concluded in Section VII, A that the intensity and capacity factors
together can describe phosphorus supply with considerable precision.
Measures of these two factors are thus required.
The simplest measurement of the intensity factor is the phosphorus
concentration in the soil solution. However, the determination of this is
complicated. For example the phosphorus concentration is affected by
the soi1:solution ratio and the ionic strength of the soil solution. As
suggested by Schofield, both complications can be reduced by standardizing the soi1:solution ratio at l : l O , using 0.01 M CaCl? as extractant.
The phosphorus in 0.01M CaCI, solution can be expressed in various
ways: ( a ) total Concentration; ( b ) concentration of individual phosphorus ions; ( c ) activity of individual phosphorus ions; ( d ) activity
products of calcium and individual phosphorus ions, e.g., % pCa
The choice of parameter will to some extent depend on the purpose
of the investigation. For plant uptake, the total concentration generally
gives a better measure than the activity (Wild, 1964), and this may be
improved by expressing it logarithmically ( Gunary and Sutton, 1967).
The capacity factor, the quantity of phosphorus that is capable of
replenishing the soil solution, can only be measured by isotopic dilution
analysis. The various methods for doing this have already been discussed in Section V, B, and the practical details for methods used in
this laboratory have been given by Gunary and Sutton (1967). They
found that of the capacity factors studied, the L value gave the best
correlations with plant uptake of phosphorus. These authors also found
that combining their best measure of the intensity factor (log P concentration) with the L value, 80 to 85 percent of the variation in phosphorus uptake by ryegrass grown in pots could be explained.
A practical method, suitable for routine laboratory analysis, which
gives a combined measure of the relevant factors, is to use an anion
exchange resin as extractant (Cooke and Hislop, 1963; Hislop and
This method causes a minimum of chemical change in the soil, and
it is well correlated with phosphorus uptake by plants.
As discussed in the preceding section the amount of phosphorus
required to maintain a particular level can be predicted if the following
are known: ( a ) the level of phosphorus required, ( b ) the rate of loss
cf labile phosphorus, ( c ) the rate of dissolution of the phosphorus
The critical level of labile phosphorus required will depend on
many agronomic factors. With present knowledge this can only be
determined initially under practical conditions using conventional field
The rate of loss of labile phosphorus may be measured in the field