Tải bản đầy đủ - 0trang
MECHANICAL RESISTANCE OF SOIL
grains ( McMurdie and Day, 1958), the elastic strain is not negligible.
Nevertheless, root growth does not seem to occur in such media, other
than in continuous pores commensurate in width with the normal, unstressed root tip. Wiersum (1957) found, for example, that the roots of
tomatoes having tips of 0.3 mm. diameter grew through sintered glass
disks in which the diameter of the pores ranged from 0.20 to 0.50 mm.,
but they were unable to penetrate disks in which the range was 0.15
to 0.20 mm.
b. Deformable matrix. Little attempt has been made to utilize the
relatively simple properties of purely cohesive media in experiments
with growing plants. Taylor and Gardner (1960) and Gardner and
Danielson (1964) measured the penetration of roots into cohesive
waxes of different hardness; these workers used the waxes simply as convenient materials with which to rank the penetrating ability of plant
roots that had been subject to various treatments or grown in various
soils ( Section VI, B, 1 ) .
Although saturated clays are often treated as purely cohesive ($I =
0 ) media in the classical theory of consolidation and bearing capacity
(Terzaghi, 1943), it is not profitable to regard them as such when dealing with deformations by plants. The reasons for this have been described
in Section 111; here we need only remind the reader that effective normal
stresses are in general not negligible around growing plant organs even
in saturated, impermeable clays.
As described in Section 11, seedlings often emerge from cohesive
soils by rupturing and lifting slabs of the overlying soil. When this kind
of deformation occurs, it is appropriate to relate emergence to the breaking forces, obtained from the dimensions of the slab and the modulus of
rupture of the soil. Relations of this kind were measured by Richards
(1953) and by Allison (1956), but their data are of little value as the
strength tests were conducted on slabs reconstituted in the laboratory
from fine earth and dried at 50°C., rather than on the crusts through
which the seedlings actually emerged. Moduli of rupture pertaining to
the moist crusts through which seedlings emerged were measured by
Hanks and Thorp (1956, 1957). But it is doubtful whether this measure
was appropriate in their experiment; emergence was not reduced by
increasing the thickness of the crust, as would have been expected if
tensile failure were the means of emergence. Rather, the shoots may have
penetrated the crust by causing continuous, local failure. On the other
hand, Parker and Taylor (1965) related the emergence of guar, Cyamopsis tetragorwloba ( L . ) Taub., to indentation test data, even though the
seedlings emerged in their experiment by rupturing a crust.
Clearly, the kind of deformation involved should be ascertained
K. P. BARLEY AND E. L. GREACEN
before mechanical criteria are chosen, Careful interpretation is needed,
particularly when one kind of deformation tends to pass into another
with time or distance. For example, radial thickening of the proximal
parts of the root sometimes ruptures peds or layers of soil. When this
happens the failure may be propagated ahead of the root apex, so that
the point stress, qp, falls to a small value. The initial penetration of the
soil by the root always involves shear failure, but, in the later stages,
penetration may result from tensile failure of the soil.
So far we have been dealing with growth in soils where compression
is unimportant, We now turn to the more general situation, where growth
is accompanied by local shear failure and compression of the soil.
Working with compressible moist clays and with gelatin, Pfeffer
(1893) related mechanical resistance, q, to the growth of radicles. He
measured q with steel probes having relieved tips. The tips were similar
in size and shape to those of the broad bean and corn radicles used in his
experiment. Initially each radicle was placed in a 2-cm. deep channel, SO
that the resistance encountered by the growing tip would be constant
during the experiment. In separate tests the channel was formed adjacent
to a glass plate so that the radicle could be observed at intervals during
elongation. When bean radicles were placed in a clay, for which the value
of q measured with the probe was 4 bar, the roots did not elongate for
the first 4 hours. After this delay elongation proceeded at a rate 20 percent below that attained in water or in a slurry of clay in water. Similar
results were obtained with corn radicles. When bean radicles were grown
in a firm gelatin ( q = 1.0 bar) elongation proceeded with little or no
delay at a rate equal to that attained in water or in a gelatin sol. Pfeffer
also reported that roots would not elongate in clays when q exceeded
12 bar, However he did not provide any experimental data for such
strong clays, and it seems likely that he merely inferred this value of q
from his knowledge of the maximum plant pressures measured in his
As the clays used in Pfeffer’s work were quite moist, gaseous diffusion
through the clay would have been very slow. In commenting on this,
Pfeffer suggested that sufficient oxygen would be obtained by diffusion
through the intercellular air space, providing the radicles were not more
than a few centimeters long. To support this argument he quoted his
observation that bean radicles grew equally well and at moderate rates
(0.8 mm.hr.-l) for at least 48 hours in either de-aired or aerated water.
Unfortunately Pfeffer did not measure the pore water pressure, u.,,,, in
his clays. In the stronger samples uw may have been large and negative,
leading to loss of turgor in the radicles (Section VI, B, 1).
Except at high void ratios a hard-grained soil is difficult to compress
RESISTANCE OF SOIL
with an isotropic pressure; nevertheless it can be compressed rather
easily when the applied stress has a large shear component. In such a soil
y is related directly to the applied pressure, p , or more generally, taking
account of pore water pressure, to the effective pressure, p’. Barley
(1963) made use of these properties to separate strength from variables
that depend upon the porosity of the medium. In his experiment a
moistened pack of 10 to 70 p Ballotini beads was housed within a modified triaxial test cell. The level of p’ in the medium could then be controlled by regulating the ambient pressure, p, applied to the pack. AS
isotropic compression was negligible ( c v = -0.002) over the range of
pressures applied ( 0 to 1 bar), volumetric air and water contents of the
bulk of the medium were not affected by the change of pressure. Corn
radicles were allowed to grow through inlets; they penetrated the medium readily, when strength permitted, by producing local shear failure.
The rate of elongation decreased from 1.6 to 1.1mm.hr.-l as p‘ increased
from 0.08 to 0.28 bar; the radicles were prevented from elongating at
p‘ = 0.58 bar. At this value of p’ the resistance offered to a cylindrical
steel probe corresponded to y = 22 bar. This is about twice as great a
pressure as the radicles are thought to be able to exert (see Table 11).
Among a number of reasons that could account for this large discrepancy
(Section 111, B, 2), differences in shape between the root tips and the
probe may be important. The problem of relating probe data to the
resistance offered to a root tip is increased, as shown in the above
experiment, by the dependence of the shape and size of the tip upon the
strength of the medium through which the root grows (see Fig. 7 ) .
Recently, three related sets of experiments have been conducted with
ordinary, unsaturated soils. Phillips and Kirkham ( 1962a) studied the
influence of mechanical resistance on the growth of corn radicles in an
unsaturated clay; Taylor and Gardner (1963) and Taylor et al., (1966)
studied both root growth and seedling emergence for a range of plant
species and soil types; Barley et al. (1965) studied the root growth of
two contrasting species in a loam. The experiments will be discussed further in Section VI, B, 2, where factor interactions are considered. Here
we are concerned with the methods used to characterize the resistance, q,
and with the values of y at which elongation growth ceased. In the last
of the investigations mentioned above point pressure, qp,and skin friction,
yf, were measured independently, and theoretical values of qp were
calculated after data had been obtained for the relevant soil parameters.
Table I11 lists the methods used in the experiments, together with the
limiting values of y. The latter have been recalculated from the original
data where necessary, and expressed uniformly as resistance per unit
cross section, y, to the penetration of a deep ( z > 3 d ) steel probe. The
The effect of mechanical resistance on the form of the elongating root tip of corn, Zea mays
L. The roots were
Values of Probe Pressure (9) a t Which Growth Ceased in Several Soil-Plant Systems
( X -1)
Phillips and Kirkham (1962a)
Taylor and Gardner (1963)
Barley et ul. (1965)
Pea and wheat
Taylor et al. (1966)
Void ratio not calculated as absolute density is not reported.
Measure of resistance
Constant load probe
Constant depth probe
constant rate probe
Constant depth probe
K. P. BARLEY AND E. L. GREACEN
wide variation found in the limiting values of q is not surprising. The low
values of q obtained by Phillips and Kirkham may arise for two reasons:
First, the samples of clay that roots failed to penetrate were nearly
saturated, and may have behaved as low 4 materials in tests with impermeable probes, Secondly, instead of allowing developed radicles to
penetrate the soil, as in the other investigations, Phillips and Kirkham
germinated the seeds within the body of cores of consolidated clay.
If pIant forces are osmotic in origin, then as x rarely exceeds 10 bar
in roots, the roots that grew through the stronger loams cannot have had
to overcome resistances as large as those opposed to the probes (see
Table 111).As we have seen in Section 111, there are a number of reasons
why root tips may meet less resistance than probes in unsaturated loams.
Briefly, the root may penetrate the soil by cylindrical rather than spherical compression; qf is likely to be less for root tips than for steel probes;
also, the propagation of cracks produced by thickening of the proximal
part of the root may reduce resistance to penetration.
It is even more difficult to relate the resistance encountered by shoots
to the limiting values of q measured by Taylor et al. As mentioned
previously, more relevant data could have been obtained if the authors
had used upward rather than downward acting probes. Apart from this,
qf may be underestimated in shallow tests or in deep tests where probes
with relieved tips are used, since the zone of maximum elongation often
lies some distance behind the apex of the shoot.
2. Heterogeneous Media
Lack of anchorage or support in loose layers of the soil may hinder
roots or shoots from penetrating stronger layers or crusts. If seed is
planted in loose soil, and the seedling shoot meets a superficial crust, the
shoot may push the seed deeper rather than emerge (Carnes, 1934).
Also, the shoot is more likely to be bent in a loose soil because of the
lack of radial support. Similarly, in a study of the ability of corn radicles
to penetrate a layer of hard wax, Taylor and Gardner (1960) found that
more radicles penetrated when grown in lightly compacted soil than
when grown in loose crumbs. In a subsequent paper Gardner and Danielson (1964) reported that compacting the soil above a wax layer failed
to improve the penetration of roots into the wax. This result appears to
be inconsistent with that obtained in 1960; but in the 1964 experiment
the hard wax was separated from the soil with a layer of soft wax, and
anchorage may have been obtained in the soft wax rather than in the soil.
Although mechanical resistance was not measured, morphological
evidence suggests that this may have been the controlling factor in an
experiment conducted by Schuurman ( 1965). His observations of the
MECHANICAL RESISTANCE OF SOIL
growth of oat roots across the boundary between layers of a humic sand
showed that roots penetrated a highly compacted layer more readily
when this was overlain by a moderately compact rather than by loose
sand. The main roots grown in loose sand branched profusely just above
the boundary with a highly compacted sand. Laterals penetrating the
highly compacted layer were short, thickened, and distorted.
Little attention has yet been given to the possible role of mechanical
factors in controlling the entry of roots into discrete peds. A recent study
of the distribution of corn roots in a silty clay loam B horizon (Edwards
et al., 1964) revealed that, although the main roots were restricted to the
voids between the peds, laterals penetrated about one-half of the total
number of discrete peds. The peds entered by the laterals were, on the
average, less dense than those which the laterals failed to penetrate; but
the factors limiting entry were not determined,
The shape and orientation of a ped in the soil may have a large influence on root entry, insofar as these factors influence the angle at which
the geotropic root tip approaches the surface of the ped. The chance of
entry is known to be reduced when the angle at which the root tip approaches a slab of hard wax becomes more acute (Gardner and Danielson, 1964). Ped size may also be important, since small peds are more
easily ruptured by internal pressure.
B. THE INTERACTION
1. The Nature of the Interaction
The growth of roots and underground shoots at a given temperature
is influenced strongly by the physical factors: mechanical resistance,
water supply, and aeration. These factors interact for two kinds of
reasons. First, as shown in Section 111, C, the factors themselves are
interdependent; secondly, the response of the plant to a change in one
factor may modify its response to another. Two examples will serve to
show the importance of this second kind of interaction.
As a soil dries uw decreases, and the plant exhibits a loss of turgor.
This is true even when transpiration is slow, as the plant and soil water
are then in or close to osmotic equilibrium. If the forces exerted by
plants arise almost entirely from osmotic turgor, as suggested in Section
IV, C, the ability of an organ to exert force should decline as uw decreases. Also, the rigidity of the root decreases as turgor is lost. Gardner
and Danielson (1964) show that the penetrating ability of plant roots
is indeed reduced for physiological reasons when u, decreases. In their
experiment cotton roots were grown through a loose soil to meet a hard
K. P. BARLEY AND E. L. GREACEN
layer of wax. The percentage of roots that penetrated the wax decreased
continuously at u, < -0.5 bar. No penetration occurred with U, <
-11 bar, Over this range of pore water pressures the water content of
the root decreased from 1200 to 600 percent, implying a considerable loss
of turgor. The authors did not determine whether the roots failed to
penetrate because they exerted less force or because they were more
easily bent. If the roots had met a firm soil rather than a wax, the effect
of u, on the resistance offered by the soil would have operated also.
The interplay of mechanical stress and oxygen supply provides a
second example of soil factor interaction mediated by plant response. In
their pressure cell experiments (Section V, A, 3), Gill and Miller ( 1956)
and Barley (1962) found that a smaller confining pressure prevented the
elongation of corn radicles when the ambient concentration of oxygen
was reduced below 5 percent. Data on root weight increase, Aw, obtained
in the latter experiment show that neither a small increase in confining
pressure ( 0.0 to 0.5 bar) nor a reduction in oxygen concentration (20 to 5
percent) reduced A w when acting singly, but, when the variabIes acted
together, A w was halved. The simplest explanation of this interaction is
provided by the effect that mechanical stress had on the shape assumed
by the radicles, the volume per unit length being twice as great when the
radicles grew under compression. This increase in bulk, taken together
with compression of the intercellular spaces, would have increased the
ambient concentration of oxygen needed to maintain a diffusive supply
of oxygen to cells in the interior of the root.
Changes in shape similar to those described above occur when roots
grow in firm media (Fig. 7 ) ; moreover, high mechanical resistance is
frequently associated with compaction and poor aeration. In a study of
the penetration of compacted soils by cotton roots, Tackett and Pearson
(1964) found that at a bulk density of 1.3 g.cc.-l roots elongated at a
high and constant rate when the oxygen concentration remained greater
than 5 percent, but at a density of 1.5 g.cc.-l the threshold concentration
of oxygen rose to 10 percent. Morphological observations suggested that
mechanical resistance was likely to have been the other factor involved.
2. Correlation of Growth with Mechanical Resistance
Although the simple correlation of mechanical properties of the soil
and plant growth may be useful in diagnosing physically adverse soil
conditions ( Culpin, 1936), correlation does not identify the particular
factors controlling growth. Fortunately, it is often possible to detect
mechanical effects by visual observation, as when shoots are seen to be
bent beneath a superficial crust. Further examples of gross morphological
symptoms are given by TayIor and Burnett ( 1964).
MECHANICAL RESISTANCE 'OF SOIL
Circumstantial evidence that mechanical resistance may be limiting is
sometimes presented by showing that other physical factors are unlikely
to be limiting. Having found a simple correlation ( r = 0.6 - 0.7) between probe resistance and corn yield in a field experiment on soil compaction, Phillips and Kirkham (1962b) proceeded to show that the soil
temperature and the oxygen content of bulk samples of the soil air were
similar at the various levels of compaction. Also, they noted that the soils
were kept moist by sprinkling during the growing period. But arguments
of this kind are often too tendentious to lead to any satisfying conclusion.
While it is possible to design experiments in which other than mechanical factors are nonlimiting, this may limit the range of values explored,
as the soil properties, e and uw,that govern air and water supply also
determine the level of mechanical resistance. A nonlimiting supply of
water can be assured by working at small pore water and osmotic pressures, and by keeping the shoots in a humid atmosphere so that transpiration is minimized. Paul (1965) found, for example, that when wheat
seedlings were covered with Mylar film to minimize transpiration, the
rate at which the roots elongated in a sand remained constant at volumetric water contents ranging from 4 to 17 percent, corresponding to
uw = -0.25 to -0.03 bar. It is more difficult to ensure a nonlimiting
supply of oxygen, particularly when mechanical properties are varied by
compaction. Various workers have employed either forced aeration
(Smith and Cook, 1946), a source of oxygen within the soil (Scott and
Erickson, 1964), or elevated ambient concentrations of oxygen (Rickman
et al., 1966). The effectiveness of such methods is best assessed by
sampling and analyzing the gas permeating the interior of the soil under
test (Tackett and Pearson, 1964). Although data on mechanical properties were not obtained in these several investigations, the results show
clearly that aeration is not the only factor involved in reducing root
growth in compact soils.
Generally we wish to examine soil-plant systems in states where each
of the physical factors capable of influencing growth may vary over a
wide range. If we aim to determine the relative importance of the different factors it becomes necessary to design experiGents to separate the
variables. To do this Phillips and Kirkham (1962a) took advantage of the
known relations between mechanical resistance, q, pore water pressure,
uw,void ratio, e, and air void ratio, e,. When uw is varied at a constant
e, q and e, change in the same sense; but when e is varied at a constant
uw, q and e, change in the opposite sense. It is true that the capillary
conductivity, k, changes in the opposite sense to q when either uw or e
is varied. However even wide changes in k may have little influence on
growth if the rate of transpiration is minimized (Paul, 1965). Also, by
K. P. BARLEY AND E. L. GREACEN
employing a soil with a high friction angle, it is not necessary to use
large negative values of u, or to vary ufowidely to obtain a useful range
The system studied by Phillips and Kirkham is described in Table I11
together with systems studied by later investigators in experiments of
similar design. Although data obtained in experiments of the kind described in Table I11 lend themselves to covariance analysis, this was
utilized only in one of the investigations (Barley et al., 1965). The partial
correlation coefficients were significant for root length, L, and resistance,
q, but not for L and e, or for L and the gaseous diffusivity parameter
( e,/e)4 proposed by Currie (1961). Though given only in graphical form,
the data of Taylor and Gardner (1963) show clearly that most of the
variation in the number of roots penetrating a sandy loam over a range
of e and u, values was associated with the variation in q. Taken together
with morphological observations, the correlative data obtained in the
experiments described in Table I11 provide unmistakable evidence that
mechanical resistance can exert a considerable influence on root growth
and seedling emergence in finely structured soils at field densities and
In the past the role of mechanical resistance has been relatively
neglected in agronomic studies, probably because of the academic separation of “soil mechanics” in schools of engineering from “soil physics”
in schools of agriculture. When applied to agronomy, engineering theories
of soil mechanics need to be modified to place more emphasis on the
compressibility of the soil, and to be combined with a knowledge of the
mechanics of plant growth.
The evidence presented in this review suggests that mechanical resistance should be regarded as having a widespread influence on the
growth of roots and underground shoots, rather than as a factor that
operates only in unusually strong soils.
Allison, L. E. 1956. Soil Sci. SOC. Am. Proc. 20, 147-151.
Amdt, W. 1965. Australian J . Soil Res. 3, 45-54.
Balla, A. 1961. Proc. 5th Intern. Conf. Soil Mech. Found. Eng. 2, 569-576.
Barley, K. P. 1954. Soil Sci. 78,205-210.
Barley, K. P. 1962. J . Exptl. Botany 13,95-110.
Barley, K. P. 1963. Soil Sci. 96, 175-180.
Barley, K. P. 1965. Australian J . Biol. Sci. 18, 499-503.
Barley, K. P., and Stolzy, L. M. 1966. Australian Soil Sci. Conf. Hbk. Paper 7-1.
Barley, K. P., Farrell, D. A., and Greacen, E. L. 1965. Australian J. Soil Res. 3, 6979.
Bell, G . G. E. 1961. J . Theoret. Biol. 1, 104-106.
MECHANICAL RESISTANCE OF SOIL
Bennett, 0. L., Ashley, D. A., and Doss, B. D. 1964. Agron. J. !%,162-165.
Bennett-Clark, T. A. 1959. In “Plant Physiology” (F. C. Stewart, ed.), Vol. 11, pp.
105-191. Academic Press, New York.
Bishop, A. W. 1960. “Pore Pressure and Suction in Soils,” pp. 3 U 6 . Butterworth,
Bishop, A. W., and Henkel, D. J. 1962. “The Measurement of Soil Properties in the
Triaxial Test.” Arnold, London.
Bishop, R. F., Hill, R., and Mott, N. F. 1945. Proc. Phys. Soc. (London) 57, 147-159.
Bordner, J. S. 1909. Botan. Gaz. 48,251-274.
Brush, W. D. 1912. Botan. Gaz. 53,453-477.
Bunning, E. 1941. Ber. Deut. Botan. Ges. 41, 2-9.
Carlson, L. 1948. Proc. 2nd Intern. Conf. Soil Mech. Found. Eng. 1, 265-270.
Carnes, A. 1934. Agr. Eng. 15, 167-171.
Childs, E. C. 1955. Proc. Natl. Acad. Sci. (India) %A, 86-92.
Croney, D., and Coleman, J. D. 1954. J. Soil Sci. 5, 75-84.
Culpin, C. 1936. J . Agr. Sci. 26, 22-34.
Currie, J. A. 1961. Brit. J. Appl. Phys. 12, 275-281.
Dainty, J. 1963. Aduan. Bot. Res. 1, 279424.
de Jong, G., and Geertsma, J. 1953. Ingenieur 65,5pp.
de Vries, H. 1879. Botan. Zh. 37, 648-654.
de Vries, H. 1884. Jahrb. Wiss. Botan. 14, 427-601.
Edwards, W. M., Fehrenbacker, J. B., and Vavra, J. P. 1964. Soil Sci. SOC. Am. Proc.
Eide, O., Hutchinson, J. N., and Landva, A. 1961. Proc. 5th Intern. Conf. Soil Mech.
Found. Eng. 2 , 4 5 5 3 .
Evans, I., and Sherratt, G. G. 1948. J. Sci. Instr. 25, 411-414.
Farrell, D. A., and Greacen, E. L. 1966. Australian J. Soil Res. 4, 1-17.
Farrell, D. A., Greacen, E. L., and Larson, W. E. 1967. Soil Sci. SOC. Am. Proc. (in
press ) .
Fountaine, E. R., and Brown, N. J. 1959. J. Agr. Eng. Res. 4, 5359.
Frey-Wyssling, A. 1952. In “Deformation and Flow in Biological Systems” (A. FreyWyssling, ed. ), pp. 194-254. North-Holland Publ., Amsterdam.
Gardner, H. R., and Danielson, R. E. 1964. Soil Sci. SOC. Am. Proc. 28, 457-461.
Gardner, W. R. 1960. Soil Sci. 89, 63-73.
Gessner, F. 1961. In “Encyclopaedia of Plant Physiology” (W. Ruhland, ed.), Vol. 16,
pp. 66&-680.Springer, Berlin.
Gill, W. R., and Miller, R. D. 1956. Soil Sci. SOC. Am. Proc. 20, 154-157.
Greacen, E. L. 1960. J. Soil Sci. 11, 313-333.
Griffith, A. A. 1924. Proc. 1 s t Intern. Congr. Appl. Math. pp. 55-63.
Haberlandt, G. 1914. “Physiological Plant Anatomy” (Transl., M. Drummond ), Macmillan, London.
Hallbauer, W. 1909. Dissertation. Leipzig. (Reviewed by 0. Damm. 1911. Botan.
Cents. 116, 201.)
Hanks, R. J., and Thorp, F. C. 1956. Soil. Sci. Soc. Am. Proc. 20, 307410.
Hanks, R. J., and Thorp, F. C. 1957. Soil Sci. SOC.Am. Proc. 21, 357359.
Hermans, J. J. 1949. In “Colloid Science” (H. R. Kruyt, ed. ), Vol. 2, p. 86. Elsevier,
Heyn, A. N. J. 1931. Records Trav. Botan. N e e d 28, 113.
Hottes, C. F. 1929. Plant Physiol. 4, 1-29.
Hvorslev, M. J. 1937. Ingenioeruidenskab. Skrifter A&, 1-159.
Kirkham, D., De Boodt, M. F., and De Leenheer, L. 1959. Soil Sci. 87, 141-144.