Mechanical impedance to root growth: a review of experimental techniques and root growth responses pot

Journalof
Soil Science,
1990,41,341-358
Mechanical impedance to root growth: a review
of
experimental techniques and root growth responses
A.
G.
BENGOUGH
&
C.
E.
MULLINS*
Cellular and Environmental Physiology Department, Scottish Crop Research Institute,
Dundee OD2 5DA and *Department
of
Plant and Soil Science, University
of
Aberdeen.
Aberdeen AB9 ?UE,
UK
SUMMARY
Mechanicalimpedancetorootgrowthisoneofthemostimportant
factorsdeterminingroot
elongation and proliferation within a soil profile. Penetrometers overestimate resistance to
root growth in soil by a factor of between two and eight and, although they remain the most
convenient method for predicting root resistance, careful interpretation
of
results and
choice of penetrometer design are essential
if

improved estimates of soil resistance to root
elongation are to be obtained. Resistance to root growth through pressurized cells contain-
ing ballotini considerably exceeds the confining pressure applied externally to these cells.
Results from this work are reappraised. Existing models of soil penetration by roots and
penetrometers are reviewed together with the factors influencing penetration resistance.
The interpretation of results from mechanical impedance experiments is examined in some
detail and root responses, including possible mechanisms of response, are discussed.
INTRODUCTION
The type of soil strength characteristic (i.e. the variation of
soil
strength with soil water content)
favourable to crop growth depends
on
both the amount and the distribution
of
the annual rainfall,
and on the nature of the crop. The soil must have sufficient mechanical strength to provide adequate
anchorage for the plant throughout its development, and to prevent the collapse of soil water and air
pathways by soil overburden pressure and the weight of vehicle and animal traffic. Dense regions
of
high strength may limit root growth and crop yield (Jamieson
et al.,
1988; Oussible, 1988) by
creating a large mechanical resistance to root growth and/or restricting the rate of oxygen supply to
roots. These dense regions occur in naturally compact soil horizons and also arise from compaction
by heavy farm machinery and by the formation of plough pans.
Mechanical impedance is experienced to varying degrees by virtually all roots growing through
soil. If continuous pores of sufficiently large diameter do not already exist, a root tip must exert a
force to deform the soil. This process may considerably decrease root elongation rates, increase the
root diameter and change the pattern of lateral root initiation (Russell, 1977).

In
this paper, the effects of mechanical impedance on root morphology are reviewed and some
direct comparisons between soil resistance
to
root growth and resistance to a penetrometer are
discussed. The physical process of root growth through soil and artificial media is considered, with
emphasis on the interpretation of results from different experimental techniques. Changes which
occur in root elongation rate under both constant, and temporally and spatially varying levels of
mechanical impedance are considered together with the complicating effects of soil aeration and
water status. Finally, possible physiological mechanisms for the root responses are discussed.
Terminology
Penetrometers provide the best estimates
of
resistance to root growth in soil, short of direct measure-
ment
of
root force. Most penetrometers consist of a metal probe with a conical tip fixed onto a
34
1
342
A.
G.
Bengough
&
C.
E.
Mullins
cylindrical shaft (Fig. 1) that is generally of smaller diameter than the cone (normally 80% of the
cone diameter; Gill, 1968; Barley
&

Greacen, 1967; Bengough, 1990). Penetrometer resistance,
Q,,
is
defined in Equation (l), where
F,
is the force required to push the penetrometer probe through the
soil, and
A,
is the cross-sectional area of the penetrometer cone:
Qp.r
=
Fp,rlAp,r
(1)
The pressure that is exerted on the soil by a growing root cannot at present be measured at every
point on the root surface. In this review, root penetration resistance,
Q,,
is defined similarly to
penetrometer resistance, but where
F,
is the component of force directed along the root-axis that the
section of root that moves through the soil must exert on the soil in order to extend, and
A,
is the root
cross-sectional area, measured behind the elongation region. In common with the literature, the
terms mechanical impedance and root penetration resistance have been used interchangeably.
-
Shaft
7
Root
hairs

41
ifl
Elongating
region
Meristemat i(
region
Phloem
Endodermis
Cortex
Epidermis
Xylem
‘‘j
,,’/
f
Muclgel
sheoth
‘l;
I
I
Fig.
1.
(a)
A
penetrometer,
where
F,,
A,,
oN.
and
a

are as defined
in
Equations
(1)
and (2).
and
(b)
a root
tip.
EFFECTS
OF
MECHANICAL IMPEDANCE ON ROOT MORPHOLOGY
When a root tip encounters an obstacle that resists penetration, the root cap becomes less pointed
and the surface cells may slough
off
(Souty, 1987). Mechanical impedance decreases the rate
of
root
elongation because
of
both a decrease in the rate of cell division in the meristem, and a decrease in
cell length (rather than volume). Eavis (1967) found a decrease of 40% in the cell division rate at a
root penetration resistance
(0.34 MPa) sufficient to decrease the root elongation rate by 70%. Cell
length is decreased and the volume
of
the inner cortical cells may decrease, but the diameter and
volume of the outer cortical and epidermal cells can be considerably greater (Barley, 1965; Wilson
et
al.,

1977). The increase in root diameter in mechanically impeded roots results mainly from an
increased thickness of the cortex; this is a consequence
of
both the increase in the diameter of the
outer cells, and an increase in the number of cells per unit length of root.
Mechanical impedance to root growth
343
The apical meristem and zone of cell extension
of
impeded roots is shorter (Barley, 1962; Souty,
1987), and root hairs develop closer to the tip of impeded roots (Goss
&
Russell, 1980). Lateral
initiation occurs nearer the tip and laterals occur together along the impeded axis (Goss
&
Russell,
1980; Barley, 1962). Where mechanical deflection causes roots to curve around an obstacle, the
initiation of laterals generally occurs on the convex side of the root (Goss
&
Russell, 1980). Root hair
development is greater on the opposite (concave) side and, in highly impeding media, the growing
zone of the root is much distorted. The growth of impeded lateral roots is affected by impedance
similarly to the main axis (Goss, 1977). However, if the pore size in the growing medium is such that
only the main root axes are impeded, the freely penetrating laterals attain much greater length than
in completely unimpeded root systems.
COMPARISON
OF
ROOT RESISTANCE WITH PENETROMETER
RESISTANCE
There have been relatively few studies involving the measurement of root force

(F,
in Equation
(1))
because of the experimental difficulties. Root force must be measured after the root has penetrated
the surface of the soil to a depth of several times its diameter (since root penetration resistance is
initially lower because the surface of the surrounding soil is displaced upwards; Gill, 1968), but
before root hairs anchor the tip (Stolzy
&
Barley, 1968; Ennos, 1989). To calculate the root
penetration resistance requires measurement of the root cross-sectional area. Root tip diameter
increases in impeded roots and, since simultaneous measurements of root diameter and force can
not normally be obtained, it is not obvious whether the initial
or
the final root diameter should be
measured. Ideally, root diameter should be recorded just behind the elongating zone and level with
the soil surface at the time of force measurement.
The results of experiments involving direct comparisons of root and penetrometer resistance
indicate that penetrometers experience two to eight times greater resistance than plant roots
penetrating soil (Table
1).
Dexter (1987) suggests that this ratio of penetrometer resistance to root
resistance is positively correlated with soil strength, being greater in ‘stronger’ soils. At present
there is neither theoretical basis for this suggestion nor sufficient published data
to
justify such
a
conclusion, although the need for accurate prediction of this ratio is clear.
Indirect evidence for the difference between root and penetrometer resistance arises from com-
paring the maximum pressures exerted by roots with penetrometer resistance in soil of sufficient
strength to virtually halt root elongation. The maximum axial pressure that a root can exert is

between about 0.9 MPa and
1.3
MPa (Misra
et
al.,
1986b), whereas root elongation stops in soil with
a penetrometer resistance of
0.8
to
5.0
MPa (Greacen
et
al.,
1969). The results are variable because of
differences between plant species and soil types, and possibly the temperatures at which the exper-
iments were performed (Greacen, 1986). Thus, roots cease elongating in soil with a penetrometer
resistance up to six
or
more times greater than the maximum axial pressure that they can exert. The
reason for this difference must be physical differences in the way in which plant roots and metal
probes penetrate soil.
MODELLING MECHANICAL IMPEDANCE TO PLANT ROOTS AND
TO
PENETROMETERS IN SOIL
Barley
&
Greacen (1967) comprehensively reviewed the mechanics
of
soil deformation and failure
which occur around penetrometer probes, roots and underground shoots. There have since been

several attempts to predict penetration resistances in soil and in ballotini beads from bulk mechan-
ical properties. All but one of these models estimate root resistance by predicting the theoretical
pressure required to expand a cavity in the soil
or
ballotini. Penetrometer resistance,
Qp,
is then given
by
Qp=a,(l
+pcota)
(Greacen
et
al.,
1968) where
uN
is the pressure required to expand a cavity in the soil (and is equal to
the normal stress on the surface of the penetrometer cone),
p
is the coefficient of soil-metal friction,
344
A.
G.
Bengough
&
C.
E.
Mullins
penetrometer probes
Table
1.

Studies involving direct measurement
of
penetration resistance both to plant roots and to
Eavis Stolzy
&
Whiteley Misra
Bengough
&
(1967)
Barley(1968)
etal.
(1981)
etal.
(1986a) Mullins(1988)
Soil
Probe diameter (mm)
Probe semiangle
Penetration rate
(mm min-
I)
mm behind tip where
root diameter
measured
ratio (orobe resistance)
remoulded remoulded remoulded
sandy loam sandy loam cores and
undisturbed
clods
of
sandy

loam
1
3
1
to2
parabolic 30" 30"
1 0.17 3
5 3t05 4
4t08 4.5 to 6 2.6 to 5.3
clay loam undisturbed
aggregates cores
of
sandy
loam
1
I
30" 30"
3 4
I*
2 to
5
1.8
to 3.8
4.5 to 9
(root resistance)
Number
of
replicates 12 2 120 324 14
*Root diameter was also measured in the air gap above the aggregate; it is
not

clear which figure was used.
and
a
is the cone semi-angle. On the assumption that plant roots experience very little frictional
resistance, Greacen
et
al.
(1968) have shown that this equation can account for much of the large
difference between the resistance experienced by plant roots and by metal probes: Equation
(2)
predicts that sharp penetrometers (i.e. small
a)
will experience a much higher component of
frictional resistance than blunter penetrometers. However, with a semi-angle of more than 30", soil
bodies (that move with the probe) have been observed to form around the probe tip
so
that soil-
metal friction is no longer involved and Equation
(2)
ceases to be applicable (Mulqeen
et
al.,
1977;
Bengough, 1988).
Farrell
&
Greacen (1966) and Greacen
et
al.
(1968) calculated the pressure required to expand

cavities in the soil by spherical and cylindrical deformation respectively. The advancing probe
or
root was accommodated by compression of the surrounding soil. This was assumed to occur in two
distinct regions: an inner zone
of
compression with plastic failure immediately surrounding the
probe, and a zone of elastic compression outside this. Sharp penetrometers
(5"
semi-angle) and plant
roots were assumed to deform the soil cylindrically, whereas blunt penetrometers caused spherical
deformation. In calculations for three sandy loam soils, the cavity pressure for cylindrical
deformation was only
25
to
40%
of that required for spherical deformation.
The major disadvantage of the Greacen
et
al.
(1968) model is that it requires many laborious
measurements of soil mechanical properties. A simpler approach was adopted by Romkens
&
Miller
(1971), who equated the pressure required for void-ratio changes occurring in a cylinder of soil
around a root with the pressure required for one-dimensional soil consolidation. The resulting
equation was used to predict the rooting densities at which further root radial expansion would be
inhibited by the expansion of neighbouring roots. Unfortunately, the Romkens
&
Miller (1971)
model is valid only for saturated cohesionless media and, therefore, is of very limited applicability to

many agricultural soils.
Further confirmation that less stress is required for radial (cylindrical) soil deformation than for
axial (spherical) deformation was provided by Abdalla
et
al.
(1969) and Hettiaratchi
&
Ferguson
(1973). For any given (elastic) strain in a cylinder of soil ahead of the root tip, it was theoretically
predicted that less stress is required to deform the soil radially than axially (Abdalla
et
al.,
1969).
Mechanical impedance
to
root growth
345
This theory was complemented by experiments using a large modified penetrometer to demonstrate
that radial expansion behind a penetrometer (or root) tip can reduce axial resistance to soil pen-
etration. Hettiaratchi
&
Ferguson (1973) predicted theoretically that the pressure required for
cylindrical soil deformation in a frictionless cohesive medium was always less than for spherical
deformation, the difference increasing with cohesion.
Collis-George
&
Yoganathan (1985) used the spherical cavity expansion model of VesiC (1972)
to
define limiting mechanical conditions for seed germination and root growth. Although this model
may be suitable to describe germination conditions, use of spherical expansion theory will have

resulted in overestimates of the resistance to root growth. The VesiC model requires fewer inputs
than the Greacen model, and includes a volumetric strain term. Collis-George
&
Yoganathan
assumed the volumetric strain to be zero,
so
that fewer soil mechanical measurements were needed
to
perform their calculation. However, their zero-strain assumption is questionable because even a
tiny volumetric strain may, under certain conditions, alter the cavity pressure considerably (VesiC,
1972).
PHYSICAL DIFFERENCES BETWEEN
SOIL
PENETRATION BY PLANT
ROOTS AND PENETROMETER PROBES
Roots are flexible organs that follow tortuous paths through the soil, apparently seeking out the
path of least resistance. They extract water from the soil, excrete mucilage from around their tips,
and swell when physically impeded. In contrast, penetrometers are rigid metal probes constrained to
a linear path through the soil. Penetrometers vary from about
0.1
mm in diameter for a small
(needle) penetrometer (e.g. Groenevelt
et al.,
1984) to over lOmm for a large (field) penetrometer
(the standard ASAE penetrometer cone has a diameter of 20.27mm; ASAE, 1969), and often
penetrate the soil at rates up to two or more orders of magnitude greater than roots (Whiteley
et al.,
1981).
The differences between penetrometers and roots have resulted in the expression of much doubt
as to the usefulness of penetrometers (e.g. Russell, 1977, p. 188), but despite their limitations they

remain the best available method of estimating resistance to root growth in soil. It is important,
therefore, to determine what are the most important physical differences between the action of roots
and penetrometers.
Rootflexibility and spatial variation
of
soil strength
Because roots often grow through cracks and holes in the soil, or follow planes of weakness between
soil peds (Russell, 1977), penetrometers are of limited use in some structured soils. Detailed work
has
been done on the behaviour of roots growing along cracks and through pores (Whiteley
&
Dexter, 1983; Dexter, 1986; Scholefield
&
Hall, 1985), but is beyond the scope of this review.
However, in coarsely structured soil, individual soil peds may be considered as continuous even
though the soil is structured on a larger scale (Greacen
et al.,
1969) and root penetration into these
peds may be important for nutrient uptake and plant growth. The forces required to buckle root tips
growing across air gaps were measured by Whiteley
&
Dexter (198
1
).
The buckling stress decreased
as the size of the air gap increased, but attempts to predict the buckling stress from the elastic
modulus of the root tip were only partly successful. Dexter (1978) has modelled root growth through
a bed
of
aggregates by relating root growth rates to penetrometer resistance within individual

aggregates, and combining this with information on the probability of roots penetrating the aggre-
gates. To date, this model has not been tested against independent experimental data over a range of
realistic conditions.
Penetrometers average soil resistance in a zone surrounding the probe tip; thus, they cannot
detect changes in soil strength that are on a scale much smaller than the tip dimensions. Groenevelt
et al.
(1984) investigated small-scale variations in strength by using a 0.15mm diameter pen-
etrometer to determine the proportion
of
linear depth in a soil core with penetrometer resistance less
than
1
MPa, and inferred that this fraction of the soil has
a
relatively low resistance to root growth.
This ‘percentage linear penetrability’ of the soil decreased at higher soil bulk density, and good
correlations between penetrability and rooting density have been obtained using a larger
(1
3 mm
346
A.
G.
Bengough
&
C.
E.
Mullins
diameter) penetrometer (Jamieson
et al.,
1988). Spectral analysis of penetrometer data, in which the

pattern
of
variation of penetrometer resistance with depth was examined using Fourier analysis, was
used by Grant
et al.
(1989, but has not yet been related to root growth.
Diameter and rate ofpenetration
Existing experimental evidence on the effects of probe
or
root diameter on penetration resistance is
based almost entirely on penetrometer measurements, and is often contradictory (Table 2).
Richards
&
Greacen (1986), in their theoretical model of cavity expansion in granular media, imply
that thin roots may deform the soil elastically, thereby encountering less resistance than thicker
roots which cause plastic deformation. However, the limited studies of several different plant species
available to date do not indicate that roots of smaller diameter are relatively less mechanically
impeded by soil or by ballotini (Gooderham, 1973; Goss, 1977). In contrast to roots, which can grow
around objects that offer high resistance to displacement, a small probe may have to displace soil
particles
of
a diameter comparable to the probe. The result is that, particularly where there is an
abundance of coarse sand
or
larger material, the effective diameter of the probe is greater than its
actual diameter
so
that smaller probes (e.g. of
1
mm rather than 2 mm diameter) can experience

a
significantly greater resistance (Whiteley
&
Dexter, 198 1).
Table
2.
Studies in which resistance to probes
of
different diameter was measured
Reference
Probe Probe type Greatest
Soil diameter (mm) (semiangle) resistance to
Dexter
&
Tanner (1973)
Barley
ef
al.
(1965)
Gooderham (1973, cited
by *below)
Bradford (1980)
Whiteley
ef
al.
(1981)*
Whiteley
&
Dexter
(1981)

Bengough (1988)
field soil
(various textures)
remoulded
sandy loam
undisturbed
undisturbed clods
and remoulded cores
of
sandy loam
remoulded
(various textures)
undisturbed cores
of
sandy loam
10,20,30,40
3.8,5.1
1.00,1.25,1.50
1.75.2.00
1.00,1.25,1.50
1.75,2.00
0.5,l
.O
sphere smallest probe
conical no difference
-
smallest probe
(30")
conical no difference
conical no difference

(307
(30")
conical smallest probe
conical smallest probe
(307
(307
It is important to distinguish between the vertical component of frictional resistance on the tip of
a penetrometer, and the friction on the shaft, which can account for a greater proportion of the total
resistance to small penetrometers (Groenevelt
et al.,
1984; Barley
et al.,
1965; Greacen, 1986).
Probes with a (relieved) shaft of smaller diameter than the tip are used to decrease the component
of
shaft friction. The success of this feature may be restricted if the trajectory of the probe tip causes the
shaft to bow and come into contact with the soil, or if soil falls or deforms inwards around the shaft
and rubs against it. Penetrometer resistance in soil cores that are laterally confined inside rigid
cylinders may also be greater if the ratio of core diameter to probe diameter is less than about
20
(Greacen
et al.,
1969). This effect
of
confinement is smaller in more compressible soils, where the
probe volume can be accommodated by the compression of a smaller cylinder of soil.
Studies on penetration rate by Eavis (1967), Voorhees
et
al.
(1975), Gerard

et al.
(1972)
and Gooderham (1973) revealed decreases of less than 20% in penetrometer resistance for decreases
in penetration rate
of
between one and three orders of magnitude down to penetration rates of
Mechanical impedance
to root
growth
347
1
mm h-’
or
slower. Although Waldron
&
Constantin (1970) found a large effect of penetration rate,
an intermittently rotated penetrometer was used, which would have resulted in larger decreases in
soil frictional resistance at slower rates of penetration. In very wet soil, penetrometer resistance is
more clearly linked to penetration rate because of its interaction with pore water pressure (Cockroft
et al.,
1969). This effect will be greater in less permeable soils (especially remoulded soil) containing a
higher propertion of silt and clay, than in sands. Penetrometer resistance doubled in
a
sandy loam
soil remoulded at approximately field capacity for a 100-fold increase in penetration rate, whereas a
250-fold increase in penetration rate resulted in only a 25% increase in penetrometer resistance in
air-dry sand (Bengough, 1988). Similarly, Cockroft
et al.
(1969) found a doubling of penetration
resistance for

a
350-fold increase in penetration rate in saturated remoulded clay. Thus, excluding
very wet and remoulded soils, penetrometer resistance is only weakly dependent on penetration rate
for speeds between those normally used in needle penetrometer measurements and typical rates of
root elongation.
Shape and friction
Root or probe shape determines both the mode of soil deformation and the amount of frictional
resistance on the tip. Observations of soil movement and density patterns surrounding probes and
roots suggest that both narrowly tapered probes and plant roots deform the soil cylindrically
compared with the spherical deformation caused by blunt probes (Cockroft
et al.,
1969; Greacen
et al.,
1968). By using both lubricated and rotated penetrometer probes it has been demonstrated
that a large component of penetrometer resistance is frictional (Tollner
&
Verma, 1984; Bengough
&
Mullins, 1988). Greacen
et al.
(1968) suggested that root tips experience virtually no frictional
resistance because of the lubricating action of mucilage secretion and the sloughing off of root cap
cells. If this is
so,
the best estimate of root resistance may be obtained by measuring the resistance
to
a
narrowly-tapered probe, and then subtracting the component of frictional resistance using
Equation
(2)

(Greacen &Oh, 1972; Voorhees
et al.,
1975).
Interactions between roots, water extraction
by
roots, root swelIing and root nutation
Because roots seldom grow through soil in complete isolation, it is important to consider inter-
actions between neighbouring
roots.
Greacen
et al.
(1969) measured resistance
to
penetration of a
narrowly-tapered probe, surrounded by six identical probes. Penetration resistance for the central
probe of a group was considerably lower than when the probe was used on its own. Tensile cracking
occurred between the probes, and similar cracks were also observed in a separate experiment
between neighbouring pea radicles growing into a loam. The drying action of roots is very important
in the formation of such cracks, which must facilitate the subsequent growth
of
lateral roots in a soil
of high resistance (Gerard
et al.,
1972). Although this cracking is generally likely to be advan-
tageous, there are soils of high tensile strength (e.g. Mullins
et al.,
1987, 1990) which may not crack
readily under the drying action of roots. In such soils, the increase in penetration resistance caused
by the soil drying may further impede root growth.
It has been suggested by Abdalla

et
al.
(1969) that roots penetrate soil by an alternating series
of
radial and axial enlargements. Graf
&
Cooke (1980) used a finite-element model and, assuming a
low coefficient of root-soil friction and that the soil behaved as a homogenous linear elastic medium,
predicted that the radial expansion of impeded root tips could reduce the axial stress on the root cap.
A
zone
of
stress relief caused by radial enlargement was also predicted by Richards
&
Greacen
(1986), and by Hettiaratchi
&
O’Callaghan (1974, 1978) who also analysed theoretically the mech-
anics of the changes in cell size and shape that can occur in mechanically impeded root tips. It is clear
that rigid penetrometers cannot mimic this radial expansion of roots, but the potential importance
of this as a mechanism for soil penetration by roots has not yet been fully investigated.
Finally, oae factor rarely commented on with respect to soil penetration is the tendency for roots
to nutate (Greacen
et al.,
1969; Ney
&
Pilet, 1981). The magnitude of this motion in highly resistant
soil is probably very small, but it could be a process by which roots locate low-resistance pathways
through heterogeneous soils, and may also reduce soil frictional resistance to root tip penetration.
Nutation has also been suggested as a mechanism that aids soil penetration by rhizomes (Fisher,

1964).
348
A.
G.
Bengough
&
C.
E.
Mullins
EFFECTS OF MECHANICAL IMPEDANCE
ON
ROOT GROWTH
Experimental techniques
Existing experimental techniques can be divided into several different catagories.
Soil.
Experiments
in
soil are more realistic, but it is difficult to ensure that mechanical impedance
is the only soil factor limiting root growth. If a soil is compacted to increase resistance to root
growth, the resulting decrease in porosity may result in poor aeration. Similarly, increasing soil-
water tension to increase soil strength may result in water stress. However, the greatest difficulty is in
determining the penetration resistance experienced by roots in soil. This is ideally determined by
direct measurement (e.g. Stolzy
&
Barley, 1968; Eavis
&
Payne, 1969; Bengough
&
Mullins, 1988),
but practical difficulties have led most researchers to use penetrometer resistance measurements to

estimate root resistance (e.g. Greacen &Oh, 1972).
Pressurizedcelis.
Root growth has been studied in artificial systems where it was intended that a
constant, uniform and known value of mechanical impedance could be imposed, while simul-
taneously maintaining a carefully controlled supply of aerated nutrient solution to the roots. Goss
(1977), Abdalla
et al.
(1969) and Barley (1963) grew roots in ballotini contained in flexible-sided cells
which could be subjected to an external confining pressure (Fig.
2).
(a)
Pressure
inlet
Root
inlet
(b)
Root inlet
Gas
inlet
Gas
under
pressure,
ud
Polythene
diaphragm
Nylon
cloth
u
Lu
024

01
2
cm
cm
Fig.
2.
(a)
A
triaxial
cell
(after Barley, 1963), and
(b)
a pressurized diaphragm apparatus (after Barley, 1962).
Fig. 2(a) is reproduced from
K.P.
Barley, Influence of
soil
strength on growth
of
roots,
Soil
Science
1963,96(3),
175-180
(0
Williams
&
Wilkins, 1963).
Resistance to root growth in ballotini has been taken as either equal to (Russell
&

Goss, 1974;
Goss, 1977;
Goss
&
Russell, 1980; Veen, 1982),
or
an order of magnitude greater than (Richards
&
Greacen, 1986; Bengough
&
Mullins, 1990) the pressure applied externally to the boundary of the
growth medium
(u3
in Fig. 2). Goss (1977), Veen (1982), Richards
&
Greacen (1986) and Bengough
&
Mullins (1990) all estimated resistance to root growth by measuring the pressure required to
inflate a small rubber tube within the pressurized ballotini cell. The tube inflation pressure depends
entirely on the detailed procedure followed (Bengough
&
Mullins, 1990).
Goss (1977), Veen (1982) and Bengough
&
Mullins (1990) inserted a tube inflated under
atmospheric pressure only and unsealed at one end into an unpressurized cell of ballotini. An
external confining pressure was then applied to the cell, causing the tube partly to deflate. The
Mechanical impedance to root growth
349
pressure required to reflate the tube to its original volume was then found to be equal to the pressure

applied externally to the ballotini cell. In contrast, when the tube is not allowed to deflate during
pressurizing of the cell, the pressure required to expand the tube beyond its initial volume is between
5
and 10 times higher than the pressure applied externally to the ballotini cell (Richards
&
Greacen,
1986; Bengough
&
Mullins, 1990). This is attributable to the frictional resistance to deformation of
the ballotini. Since roots penetrating the ballotini must exert pressure to expand a new cavity in
previously undisturbed ballotini, the latter experiment gives more accurate representation
of
the
resistance experienced by growing roots. Although the pressure required to inflate a tube in the
ballotini will depend on both the tube diameter and on the frictional properties of the tube walls, it
seems reasonable to conclude that the resistance to root growth in the cells considerably exceeded
the external confining pressure.
Richards
&
Greacen (1986) also used a finite-element model
to
predict the effect
of
tube inflation
pressure on tube diameter in ballotini and in sand. Tube diameter increased slowly until the internal
pressure reached a certain critical value, when the diameter increased much more rapidly (Fig.
3).
Predicted inflation pressures were much higher than the external confining pressure, and the
difference was greater
for

sand than
for
ballotini.
E
E
v
I
I
I
I
I
I
I
I
i
I
I
I
i
I
I
3.0}
I
/
I
36~
I
I
I
I

I
I
I
/
/
/
I
1
I
1
I
I
0
0.05
0.10
0.15
0.20
0.25
Inflation
pressure
(MPa)
Fig.
3.
Outside diameter
of
tube
vs
inflation pressure in ballotini at several constant external cell pressures
(indicated at the top
of

each curve), after Richards
&
Greacen
(1986).
Experimental
(-);
theoretical
( )
Pressurized diaphragms.
Barley (1962) and Gill
&
Miller (1956) adopted slightly different tech-
niques, using an externally-pressurized rubber diaphragm to exert pressure on roots growing down a
porous plate, separated from the diaphragm by nylon cloth
or
ballotini respectively (Fig.
2).
In these
experiments, asymmetric stress was applied to the whole root length. These authors suggested that
resistance to root growth might be considerably greater than the diaphragm pressure
(o,, in Fig.
2).
Barley (1962) estimated resistance to root growth by measuring the pressure required to inflate a
small tube placed under the diaphragm. He assumed a coefficient of soil-root friction of
0.25
and
estimated that resistance to root elongation varied between
2
and
7

times the diaphragm pressure.
Gill
&
Miller (1956) observed that if roots were well-covered with ballotini, elongation ceased at
lower diaphragm pressures. The authors suggested that ‘arching’ of the ballotini caused small
displacements of single glass beads to require larger displacements of the diaphragm. Thus,
resistance to root elongation was considerably greater than the diaphragm pressure.
Pressurized airlwater.
Chaudhary
&
Aggarwal(l984) proposed growing seedlings in moist sand
inside a pressure vessel as an experimental technique to measure the effect of mechanical resistance
on root elongation. However, although application of air pressure would cause a corresponding
350
A.
G.
Bengough
&
C.
E.
Mullins
increase in the absolute value of root cell turgor pressure, the decrease in osmotic potential or cell
wall tension required for root cell extension would
be
independent of the externally applied pressure
(assuming the cell permeability to water remained unchanged). Thus, the experiment does not truly
represent the situation of roots elongating against an external mechanical resistance. The reason for
the large decrease in root growth rates observed by the authors could be the 10-fold increase in
dissolved gas concentration in the water inside the pressure chamber, and ultimately in the plant,
resulting from the 10-fold increase in air pressure that they applied (Henry's law).

Effect
of
a constant mechanical impedance
on
root
elongation rate
Plots of relative root elongation rate against pressure are shown in Figs 4 and
5,
and the techniques
used are summarized in Table 3. Root elongation rate varies approximately inversely with increas-
ing soil penetrometer resistance (Fig. 4). Because penetrometers experience a resistance
2
to 8 times
greater than that experienced by roots, the maximum penetrometer resistance measured in a
medium when roots can only just elongate far exceeds the maximum stress which roots can exert
(about 0.9 to 1.3 MPa). In contrast, results from studies in artificial systems (Fig.
5)
show root
elongation virtually halted by pressures applied to the outside of ballotini-filled cells of less than
0.1
MPa
(Goss,
1977; Abdalla
et al.,
1969; Barley, 1963). This is only one-tenth
of
the maximum
stress that roots can exert and, as already explained, is due to the resistance to root growth being
considerably greater than the externally applied pressure.
m m

?
5;
60
t
m
W
3
0
-
c
0

5
20
c
0
W
-
Penetrometer resistance (MPa)
Fig.
4.
Root elongation
vs
penetrometer resistance obtained by Voorhees et
al.,
1975
(A
=
sandy loam,
B

=clay)
and by Taylor
&
Ratliff, 1969
(C
=
peanuts, D =cotton). Details are summarized in Table
3.
There have been only three studies in which both root elongation rate and root resistance (as
directly determined from the root force) have been measured (Eavis, 1967; Stolzy
&
Barley, 1968;
Bengough
&
Mullins, 1988). The results of Eavis (1967) have been replotted in Fig.
5
to show the
relationship between root elongation rate and root penetration resistance. This shows that root
elongation rate was reduced by about
50%
by a resistance to root growth
of
approximately
0.3 MPa.
Bengough
&
Mullins (1988) found that maize root elongation rates in cores of two sandy loam
soils were reduced to between
50%
and 90% of that of control plants grown in loose sieved soils, by

root penetration resistances of between 0.39 and 0.48 MPa (based on the initial root tip cross-
sectional area). Stolzy
&
Barley (1968) found that the elongation rates
of
two pea radicles were
reduced to
44%
of
their unimpeded rate by a root penetration resistance of 0.46 MPa.
Mechanical impedance
to root growth
35
1
100
-
+
c
E
+
0
80
P
+
ul ul
P
60
c
0
._

20
c
0
-
W
0
0.
I
0.2
0.3
0.4
0.5
Stress
(MPa)
Fig.
5.
Root elongation
vs
stress, as specified below (see Table
3
for
details) obtained in six studies:
A.
Abdalla
et
al.
(1969): external confining pressure
9.
Barley (1962): estimated root penetration resistance
C.

Barley (1963): external confining pressure
D.
Eavis (1967): measured root penetration resistance
E.
Gill
&
Miller (1956): diaphragm pressure
F.
Goss (1977): external confining pressure
G.
Greacen
&
Oh
(1972): normal stress on penetrometer cone.
The results of these direct experiments represent the maximum effect
of
mechanical impedance
to be expected on root elongation rate for any particular root resistance. Reduced oxygen or
nutrient supply to the roots can further limit the observed 'root elongation rate, and similarly
anchorage of the tip by root hairs can result in the resistance being underestimated.
All
the experimental relationships in Figs
4
and
5
show similar patterns in which elongation rate
decreases with increasing mechanical resistance.
If
the results of direct experimental measurements
of root resistance and elongation rate (Stolzy

&
Barley, 1968; Eavis, 1967) are compared with results
from penetrometer studies (e.g. Taylor
&
Ratliff, 1969), the factor of between
2
and 8 by which
penetrometer resistance exceeds root resistance is sufficient to remove any apparent discrepancy.
Similarly, the factor of between
5
and 10 by which resistance to root growth exceeds the confining
pressure externally applied to the boundary of the growth medium is sufficient to bring the results
of
experiments performed in ballotini cells into broad agreement with the other experimental
approaches.
Root elongation in relation to resistance
which
varies spatially or temporally
Spatial variation in root penetration resistance occurs when roots enter
or
leave voids
or
soil
peds.
Temporal variation may also arise if the soil matric potential around the root tip changes, for
example owing to water extraction by roots or rainwater infiltration. Despite its practical import-
ance, relatively little work has been done on the transitions which occur in root elongation rate when
the resistance to root growth changes.
Spatial variation.
The elongation rate of

roots
growing into
a
zone
of
greater mechanical impe-
dance has been observed to decrease with time (Barley, 1962, 1963). However, Barley notes that
352
A.
G.
Bengough
&
C.
E.
Mullins
Table
3. Techniques used in studies of the effect of mechanical impedance on root elongation
rate (Figs
4
and
5)
Plant Root elongation
Reference species Root growth medium plotted against
Abdalla
e?
a/.
(1969)
Barley
(1962)
Barley

(1963)
Eavis
(1967)
Ehlers
et
a/.
(1983)
Gill
&
Miller
(1956)
Goss
(1977)
Greacen
&
Oh
(1972)
Taylor
&
Ratliff
(1
969)
Voorhees
et
a/.
(1
975)
barley
maize
maize

peas
oats
maize
barley
cotton and
peanuts
peas
200
pm ballotini
in cell subject to a
confining pressure
porous plate under
diaphragm covered by
a nylon cloth
1C70
pm ballotini
in cell subject to
a confining pressure
remoulded
2
mm sieved
sandy loam soil at
different bulk
densities and matric
potentials
tilled and no-tilled
silt loam field soil
50
pm ballotini between
porous plate and rubber

diaphragm
100
pm ballotini in
cell subject to a
confining pressure
remoulded sandy loam
soil at different bulk
densities and matric
potentials
remoulded
2
mm sieved
loamy sand at different
bulk densities and
matric potentials
remoulded
2
mm sieved
sandy loam and clay at
-0.33
MPa matric
potential
external
confining pressure
estimated
resistance to root
growth
external confining
pressure
root resistance

measured directly
penetrometer
resistance
(1
1 mm
diameter,
30"
semi-
angle probe)
pressure on
diaphragm
external confining
pressure
penetrometer
resistance after
subtracting
friction
(5"
semi-
angle)
penetrometer
resistance
(3.2
mm
diameter,
30"
semi-
angle probe)
penetrometer
resistance

(1.77
mm
diameter,
30"
semi-angle probe)
the effects may have merely been due to a greater length of root being compressed (Barley,
1962),
and resistance to root growth increasing gradually with depth (Barley,
1963),
respectively. The
elongation rate of roots encountering a very small mechanical resistance has been found to decrease
transiently by a surprisingly large amount
(Goss
&
Russell,
1980).
Maize root elongation rates
decreased to less than one-third of their unimpeded rate during the
10
min period after starting to
grow through loosely packed ballotini. However, during the subsequent 10 min the elongation rate
Mechanical impedance
to
root
growth
353
returned to its original unimpeded value. Identical experiments performed using roots that first had
their root caps removed showed no significant changes in root growth rate. Thus, the root cap must
have a role in regulating root response to very low resistances, although it is uncertain whether the
same mechanism operates

at
the greater resistances more often found in soil.
Temporal variation.
Mechanical pressure applied to the apical 15 mm of previously unimpeded
roots can halt root elongation which resumes only after a lag time (13 to 50 h), which itself increases
with increasing stress (Barley, 1962). Localized moderate pressure (0.1 MPa) applied to the root
apex results in a bigger reduction in elongation rate than when fully enlarged tissue is compressed
(Barley, 1965). Higher pressures applied to the older tissue caused browning,
loss
of turgor and
stopped elongation, although no tissue damage resulted if the root cells were allowed
to
enlarge and
mature under the mechanical pressure (Barley, 1965). When the external confining pressure is
reduced to zero, mechanically impeded roots growing in pressurized cells of ballotini only gradually
return to the unimpeded elongation rate (Goss
&
Russell, 1980). The delay, of 2 to 7 d, was longer for
the more severely impeded roots. Thus, roots which develop under mechanical impedance probably
undergo physiological changes to adapt to the stress.
Interaction
of
mechanical impedance with matric potential and aeration
Roots grow more slowly in poorly aerated soil (Blackwell
&
Wells, 1983; Greenwood, 1969) and in
soil of low (i.e. more negative) matric potential (Eavis
&
Payne, 1969; Eavis, 1972; Yappa
et

al.,
1988): but do these two stresses interact with mechanical impedance to produce an effect that is
greater than would be expected from each stress acting independently?
Gill
&
Miller (1956) and Barley (1962) grew maize roots in ballotini compressed by a diaphragm
and circulated with nutrient solution containing different concentrations of oxygen. Gill
&
Miller
(1956) found that root elongation ceased when the oxygen concentration was reduced to 1% in
mechanically impeded roots at an externally applied diaphragm pressure of
0.15
MPa or greater,
whereas growth continued in the better-aerated treatments. Barley (1962) found that increasing the
externally applied diaphragm pressure from zero to
0.05
MPa resulted in a 31% decrease in root
elongation rate for roots grown in solution containing a 20% oxygen concentration, whereas a
50%
reduction in root elongation rate resulted in roots grown in solution containing a 5% oxygen
concentration. Goss
et
al.
(1989) found a small interaction of mechanical impedance with aeration
in
maize roots, but the interaction was not statistically significant in wheat. The explanation for the
interaction between oxygen deficiency and mechanical impedance is still uncertain, but may be due
to
mechanical impedance changing the morphology of the cortical root cells and the spaces between
these cells. This might result in a shorter and a more tortuous path for oxygen transport along the

root to the tip, and for ethylene transport away from the tip.
It
is more difficult to ascertain whether low matric potentials interact with mechanical
impedance, because soil strength and root penetration resistance increase as the matric potential
decreases. Taylor
&
Ratliff (1969) measured the root elongation rates of cotton and peanuts in
remoulded soil at several different bulk densities and matric potentials in the range -17 to
-700 kPa for cotton, and
-
19 to
-
1250 kPa for peanuts. Root elongation rate clearly depended
on the penetrometer resistance of the soil and not on the matric potential
per se.
Similar results
were obtained by Greacen
&
Oh (1972) for peas at matric potentials below -0.5 MPa, and by
Taylor
&
Gardner (1963) for cotton root penetration at matric potentials between
-
20 kPa and
-
67 kPa. These results imply that there is no interaction between matric potential and mechanical
impedance.
In contrast to these findings, Mirreh
&
Ketcheson (1972) found that, in soil with penetrometer

resistance in the range
0
to
2.5
MPa, the root elongation rate of maize was progressively reduced for
any given penetrometer resistance as matric potential was decreased from -0.1 to -0.8 MPa. The
reduction in elongation rate was greatest in soil with high penetrometer resistance, suggesting an
interaction between matric potential and mechanical impedance.
A
source of uncertainty in these experiments is that the matric potential occurring close to the
root surface may be considerably lower than that in the bulk of the soil, and will decrease as the
354
A.
G.
Bengough
&
C.
E.
Mullins
transpirational demand increases (Cowan, 1965; Hainsworth
&
Aylmore, 1986). Thus, roots grow-
ing in soils of differing bulk matric potential may experience similar matric potentials at the root
surface (Mirreh
&
Ketcheson, 1972).
Possible
mechanisms
for
the

response
of
roots
to mechanical impedance
The growth rate of root cells is controlled by the balance of the internal and external pressures
together with the rheological properties of the cell wall. Lockhart’s (1965) equation for cell growth
was modified by Greacen (1986) to allow for root penetration resistance
gr
(Pa) giving,
dlldt
=
ZKrlm(n
-
W
-
0,)/(2K
+
r’m)
(3)
where dl/dt is the rate of increase in cell length (m
s-’),
K(m Pa-ls-’) is the water permeability
of
the
cell membrane per unit surface area, m (m-’Pa-Is-’) is the cell wall extensibility,
n
(Pa) is the
osmotic pressure,
r
(m) is the cell radius and W(Pa) is the yield stress of the cell wall. Greacen (1986)

used a simplified version of the above equation together with Greacen
&
Oh’s (1972) model of
osmotic adjustment to generate a relationship between soil resistance and root elongation rate. The
calculations produced a curve similar to that found by Eavis (1967) and Greacen
&
Oh (1972) by
assuming a progressive decrease in the ratio of cell length to diameter, and also a linear reduction in
cell production with increase in root penetration resistance.
Greacen
&
Oh (1972) proposed that root cells adjust their osmotic potential (the term ‘osmo-
regulation’ has recently been criticized as misleading; Munns, 1988; Reed, 1984) to compensate for
increased soil resistance and for low soil-water potential. The efficiency
of
this osmotic adjustment
was measured to be 70%
for
mechanical impedance and 100% for water potentials down to
-0.8
MPa. The agreement
of
their results with such an ‘efficiency’ hypothesis was questioned by
Russell
&
Goss (1974) although Greacen (1986) argued that any discrepancy was caused by the high
degree of variability associated with this aspect
of
root behaviour. Atwell (1988) recently suggested
that the decrease in osmotic potential observed in mechanically impeded root tips occurred because

of a simple build-up of solutes because
of
the slower root growth rate. Such an excess of solute
supply over demand has also been suggested as the source of osmotic adjustment in plants growing
in dry or saline soils (Munns, 1988).
However, a model of root response to mechanical impedance based entirely on osmotic adjust-
ment does not explain the many morphological changes that occur in mechanically impeded roots.
Barley (1976) and Feldman (1984) suggested that ethylene may be involved in root response to
mechanical impedance. Kays
et
al.
(1974) noted that the rate of ethylene evolution increased by up
to six times when bean roots were impeded by a barrier. Exogenous applications
of
ethylene can
cause root radial thickening, a reduction in the root elongation rate, and a profusion
of
root hairs
within
1
to 2 mm of the swollen root apices (Kays
et al.,
1974; Smith
&
Robertson, 197
I);
all of these
have been observed in mechanically impeded roots.
Cell shape can be determined by the orientation of the cellulose microfibrils as they are laid down
in the cell wall (Osborne, 1976). Initially these are aligned transversely to the root axis, but as cellular

elongation begins to slow, the orientation of the microfibrils becomes more oblique (Hogetsu, 1986;
Barlow, 1989). Veen (1982) observed longitudinal deposition of microfibrils in mechanically
impeded roots, which resulted in shorter fatter cells. Such longitudinal deposition also occurs in
roots exposed to ethylene (Osborne, 1976). The smaller diameter of the inner compared with the
outer cortical cells, observed in impeded roots by Wilson
et al.
(1977), may have resulted from the
innermost cell layers experiencing considerable wall pressure from the outer cells (Veen, 1982).
Thus, despite the presence of longitudinal microfibrils, lateral expansion of the innermost cells could
not occur.
Dexter (1987) presented
a
physical model of root growth in terms of the balance of pressures
acting on the root, based largely on reworking published results using previously derived empirical
formulae. An implicit assumption of this model is that the cell wall ‘extensibility’ is independent
of
the mechanical impedance experienced by the root. Surprisingly he does not comment on this
assumption, although the evidence on microfibril deposition already presented suggests that cell
wall extensibility may be restricted in the longitudinal direction. 3,5-diiodo-4-hydroxy benzoic acid
Mechanical impedance to root growth
355
(DIHB) is thought to be antagonistic to endogenous ethylene production/action, and increases the
elongation rate of mechanically impeded roots, possibly by influencing the action of ethylene on cell
wall extensibility
(Goss
et
al.,
1987).
Recently the involvement of ethylene in root response to mechanical impedance has been
seriously questioned (Moss

et
al.,
1988). Moss et
al.
measured a two- to 2.5-fold increase in the rate
of ethylene evolution by mechanically impeded maize roots, and found that supplying ethylene in
the air-flow to unimpeded roots simulated closely the effects of impedence. However, two inhibitors
of ethylene production,
aminoethoxyvinylglycine,
and action, 2,5-norbornadiene, which overcame
the effects of ethylene in unimpeded roots, did not modify the growth
of
mechanically impeded
roots, although the increased rate of ethylene evolution was entirely suppressed.
Moss
et
al.
(1988) suggest that the increased rate
of
ethylene evolution by impeded roots may not
be a result of a direct effect of mechanical impedance, but of the physical wounding of the radially
expanding root by the ballotini (such self-inflicted damage has been observed during time-lapse
studies of bean root growth in ballotini; personal observation, AGB). However, the wounding
hypothesis does not entirely explain why an increase in the rate
of
ethylene evolution occurred
within
1
h ofwhen bean roots grown inside perforated tubes were mechanically impeded by a plastic
barrier (Kays

et
al.,
1974). The likelihood of physical wounding of the root tissue in this system
should be considerably smaller than in the ballotini, although it is possible that some damage to the
root tip occurred as it buckled upon reaching the barrier. Clearly the role of ethylene in the response
of roots to mechanical impedance must be re-examined, as much of the supporting evidence for its
involvement is largely correlative.
CONCLUSIONS
Except where direct measurements of root force are available, penetrometer resistance is currently
the best method for estimating resistance to root growth in soil, although subject to many limi-
tations. The relationship between penetrometer resistance and root resistance needs to be more
clearly established by direct measurements of both root force and penetrometer resistance in a
variety of differing soils over a range of bulk densities and matric potentials. It is probable that
improved estimates of root resistance may be obtained by subtracting the frictional component of
penetrometer resistance (determined by independent measurement
of
the coefficient
of
soil-metal
friction) from the total resistance to a narrowly tapered probe. The effects on penetration resistance
of
penetrometer or root diameter still require a fuller investigation.
Root elongation rate is progressively decreased by increasing mechanical resistance to growth,
and ceases at root penetration resistances of about
1
MPa. Resistance to root elongation within a
ballotini cell confined by an externally applied pressure may be up to one order of magnitude greater
than the externally applied pressure. The mistaken interpretation
of
results from such experiments

has caused the apparent disagreement with results from penetrometer studies in soil. Pressurized
ballotini cells still provide a valuable technique for studying the effects
of
mechanical impedence on
root morphology and nutrient uptake (Goss, 1977; Lindberg
&
Pettersson, 1985), and are poten-
tially
of
considerable use for investigating the largely unresearched ability
of
different species and
varieties to grow against mechanical impedance (Goss, 1974). The changes in root growth rate
which occur as a root enters or leaves a zone
of
high mechanical impedance are of considerable
relevance to real soils, but have received surprisingly little attention.
The mechanism
of
root response to mechanical impedance may involve osmotic adjustment,
although insufficient evidence has been presented to test such a model. The role of ethylene must be
re-assessed in the light of some recent experiments, and its type
of
involvement in the response
of
roots to mechanical impedance must be more clearly established.
ACKNOWLEDGEMENTS
We thank Drs D. Linehan, B. Marshall,
I.
Young, D. Robinson and other colleagues at SCRI for

their helpful discussions and constructive criticism of this manuscript.
356
A.
G.
Bengough
&
C.
E.
Mullins
REFERENCES
ABDALLA, A.M., HETTIARATCHI, D.R.P.
&
REECE,
A.R. 1969. The mechanics of root growth in granu-
lar media.
Journal of Agricultural Engineering
Research
14,263-248.
ASAE 1969. ASAE Recommendation: ASAE R313:
Soil Cone Penetrometer. American Society of
Agricultural Engineers, St. Joseph, MI.
ATWELL, B.J. 1988. Physiological response of lupin
roots to soil compaction.
Plant and Soil
111,
BARLEY, K.P. 1962. Theeffects ofmechanical stresson
the growth
of
roots.
Journal of Experimental Botany

13,95-110.
BARLEY, K.P. 1963. Influence of soil strength on
growth of roots.
Soil Science
96,175-180.
BARLEY, K.P. 1965. The effect oflocalized pressure on
the growth of the maize radicle.
Australian Journal
of Biological Science
18,499-503.
BARLEY, K.P. 1976. Mechanical resistance of the soil
in relation to the growth of roots and emerging
shoots.
Agrochimica
20,173-182.
BARLEY, K.P.
&
GREACEN, E.L. 1967. Mechanical
resistance as a soil factor influencing the growth
of roots and underground shoots.
Advances in
Agronomy
19.1-40.
BARLEY, K.P., GREACEN, E.L.
&
FARRELL,
D.A. 1965.
The influence of soil strength on the penetration
of
a loam by plant roots.

Australian Journal of Soil
Research
3,69-79.
BARLOW, P.W. 1989. Anatomical controls of root
growth.
Aspects of Applied Biology
22,57-66.
BENGOUGH, A.G. 1988. The
use
of penetrometers in
estimating mechanical impedance to root growth.
Ph.D. thesis, University of Aberdeen.
BENGOUGH, A.G. 1990. The penetrometer in relation
to mechanical resistance to root growth. In
Soil
Analysis: Physical Methods
(eds K.A. Smith
&
C.E.
Mullins), pp. 431435. Marcel Dekker, New York.
BENGOUGH, A.G.
&
MULLINS, C.E. 1988.
Use
of a low-
friction penetrometer to estimate soil resistance to
root growth.
Proceedings of the
11th
Conference of

the International Soil Tillage Research Organisation
BENCOUGH, A.G.
&
MULLINS, C.E. 1990. The resist-
ance experienced by roots growing in a pressurised
cell-a reappraisal.
Plant and Soil
123.73-82.
BLACKWELL,
P.S.
&
WELLS, E.A. 1983. Limiting oxy-
gen flux densities for oat root extension.
Plant and
Soil73,129-139.
BRADFORD, J.M. 1980. The penetration resistance
of
a
soil with well defined structural units.
Soil Science
Society of American
Journal44,601-606.
CHAUDHARY, M.R.
&
AGGARWAL, G.C. 1984. A
simple technique to evaluate the effect of mechan-
ical stress on root growth.
Journal of Agricultural
Science
102,79-80.

COCKROFT, B., BARLEY, K.P.
&
GREACEN, E.L. 1969.
The penetration of clays by fine probes and root
tips.
Australian Journalof Soil Research
7,333-348.
COLLIS-GEORGE,
N.
&
YOGANATHAN, P. 1985. The
effect of
soil
strength
on
germination and emerg-
ence of wheat
(Triticum aestivum
L.).
11.
High shear
strength conditions.
Australian Journal
of
Soil
Research
23,589-601.
277-28
1.
1, 1-6.

COWAN, I.R. 1965. Transport of water in the
soil-
plant-atmosphere system.
Journal of Applied
DEXTER, A.R. 1978. A stochasticmodel for the growth
of roots in tilled soils.
Journal of Soil Science
29,
DEXTER, A.R. 1986. Model experiments
on the behav-
iour of roots at the interface between a tilled seed-
bed and a compacted sub-soil. 11. Entry of pea and
wheat roots into subsoil cracks.
Plant and Soil
95,
DEXTER, A.R. 1987. Mechanics of root growth.
Hunt
andSoil98,303-312.
DEXTER,
A.R.
&
TANNER, D.W. 1973. The force on
spheres penetrating soil.
Journal of Terramechanics
EAVIS, B.W. 1967. Mechanical impedance to root
growth. Agricultural Engineering Symposium,
Silsoe. Paper 4/F/39, pp.
1-1
1.
EAVIS, B.W. 1972. Soil physical conditions affecting

seedling root growth. I. Mechanical impedance.
aeration and moisture availability as influenced by
bulk density and moisture levels in a sandy
loam
soil.
Plant and
Soil36,613-622.
EAVIS, B.W.
&
PAYNE, D. 1969. Soil physical con-
ditions and root growth. In
Root Growth
(ed. W.J.
Whittington), pp. 315-336. Butterworths, London.
EHLERS, W., KOPKE, U., HESSE,
F.
and BOHM, W. 1983.
Penetration resistance and root growth of oats in
tilled and untilled Loess soil.
Soil and Tillage
Research
3,261-275.
ENNOS, R. 1989. The mechanics of anchorage in seed-
lings of sunflower,
Helianthus annuus
L.
New
Phyrologist
113, 185-192.
FARRELL, D.A.

&
GREACEN, E.L. 1966. Resistance to
penetration of fine probes in compressible soil.
Australian Journal
of
Soil Research
4,
1-1
7.
FELDMAN, L.J. 1984. Regulation of root development.
Annual Review of Plant Physiology
35,223-242.
FISHER, J.E. 1984. Evidence of circumnutational
growth movements of rhizomes of
Poapratensis
L.
that aid in soil penetration.
Canadian Journal
of
Botany
42,293-299.
GERARD, C.J., MEHTA, H.C.
&
HINOJOSA, E. 1972.
Root growth in a clay soil.
Soil Science
114,37-49.
GILL, W.R. 1968. Influence of compaction hardening
of soil
on

penetration resistance.
Transactions of
the American Society of Agricultural Engineers
11,
GILL, W.R. &MILLER,
R.D.
1956. A method for study
of the influence of mechanical impedance and
aeration on the growth ofseedling roots.
SoilScience
Society of America Proceedings
20, 154-1 57.
GOODERHAM, P.T. 1973. Soil physical conditions and
plant growth. Ph.D. Thesis, University of Reading.
GOB,
M.J
1974. Effects
of
mechanical impedance on
root growth. Ph.D. Thesis, University of Reading.
Goss, M.J. 1977. Effects of mechanical impedance
on root growth in barley
(Hordeum vulgare
L.).
I.
Effects
on
elongation and branching of seminal
roots.
Journal of Experimental Botany

28,96-11
I.
Goss,
M.J.
&
RUSSELL, R.S. 1980. Effects of mechan-
ical impedance on root growth in barley
(Hordeum
vulgare
L.).
111. Observations on the mechanism
Ecology
2,221-239.
102-1 16.
135-147.
9,31-39.
74 1-753.
Mechanical impedance to root growth
357
of
response.
Journal
of
Experimental Botany
31,
577-588.
Goss,
M.J., BARRACLOUGH, P.B.
&
POWELL, B.A.

1989. The extent to which physical factors in the
rooting zone limit crop growth.
Aspects
of
Applied
Biology22,173-183.
Go%
M.J., DEXTER, A.R.
&
EVANS, M. 1987.
Mechanics of root elongation and the effects of3,5-
diiodo-4-hydroxybenzoic acid (DIHB).
Plant and
S0il99,211-218.
GRAF,
G.L.
&COOKE, J.R. 1980. Soil stressesaround a
growing root tip.
Proceedings
of
the American
Society
of
Agricultural Engineers:
80-1024.
GRANT, C.D., KAY, B.D., GROENEVELT, P.H.9 KIDD,
G.E.
&
THURTELL, G.W. 1985. Spectral analysis
of

micropenetrometer data to characterize soil struc-
ture.
Canadian Journal
ofsoil
Science
65,789-804.
GREACEN,
E.L. 1986. Root response to soilmechanical
properties.
Transactions
of
the 13th Congress
of
the
International Society
of
Soil Science
5,20-47.
GREACEN, E.L. OH, J.s. 1972.
Physics
Of
root
growth.
Mature, New Biology
235,24-25.
The mechanics of root growth in soils with particu-
lar reference to the implications for root distri-
pp. 256-268. Butterworths, London.
LOCKHART, J.A. 1965. An analysis of irreversible
cell elongation.

Journal
of
Theoretical Bioiogy
8,
264-275.
MIRREH, H.F.
&
KETCHESON, J.W. 1972. Influence of
soil water matric potential on corn root elongation.
Canadian Journal
of
Soil Science
53,383-388.
MISRA, R.K., DEXTER, A.R. &ALSTON, A.M. l986a.
Penetration ofsoil aggregates of finite size. 11. Plant
roots.
Plant and Soil
94,59-85.
MISRA, R.K., DEXTER, A.R.
&
ALSTON, A.M. 1986b.
Maximum axial and radial growth pressures of
plant roots,
Plant
andSoil95,315-326.
Moss,
G.I.,
HALL, K.C. and JACKSON, M.B. 1988.
Ethylene and the responses of roots of maize
(Zea

mays
L.) to physical impedance.
New Phytologist
109,303-31
1,
M~~~~~~,
c,E,,
yOUNG, I,M,, BEN~UGH,
A,G,
&
L~~,
G.J,
1987, Hard setting soils,
soit
use
and
M~~~~~~~~~
3,
79-83,
MULLINS, C.E., MACLEOD, D.A., NORTHCOTE, K.H.,
TISDALL, J.M.
&
YOUNG, I.M. 1990. Hardsetting
soils: Behaviour, occurrence and managment.
M~~~~~~~,
J,,
sTAFFORD,
J.V.
&T~~~~~,
D,w. 1977,

Evaluation
of
penetrometers
for
measuring
soil
MUNNS, R. 1988. Why measure osmotic adjustment?
Australian Journal
of
Plant Physiology
15,717-726.
georeacting
roots,
and
Environment
4,
339-343.
OSBORNE, D.J. 1976. Control ofcell shape and cell size
Perspectives in Experimental Biology,
Vol.
2 (ed.
N'
ousslBLE, M. 1988. Effect Of subsoiling
a
compacted
clay loam soil on the nitrogen uptake ofwheat.
Pro-
ceedings
Of
the

'Ith
Conference
Of
lhe
International
GREACEN,
E.L., BARLEY, K.P.
&
FARRELL,
D.A. 1969.
Advances
in
~~il~~i~~~~
11,37-107.
bution. In
Root
Growth
(ed.
w.J.
Whittington),
strength,
Journal
of
Terramechanics
14, 137-1
51,
GREACEN3
E'L', FARRELL3
D'A'
'

CoCKRoFT,
B'
1968.
Transactions
of
the 9th Congress
of
the International
Society
of
Soil
Science
1,769-779.
GREENWOOD, D.J. 1969. Effect of oxygen distribution
in
the
soil
on
plant
growth,
In
Root
Growth
(ed,
London.
'Oil
resistance
to
probes
and

plant
roots'
NE~, D,
&
PILET, P.E, 1981. Nutation ofgrowing and
W,J, Whittington), pp, 202-221. Butterworths,
by
the
regulation Of
auxin
and
ethylene' In
GROENEVELT, P.H., KAY, B.D.
&
GRANT, C.D. 1984.
pp'
89-102'
'ergamon'
Oxford'
physical assessment
ofa
with respect to rooting
HAINSWORTH, J.M.
&
AYLMOE, L.A,G, 1986, Water
extraction by single plant roots,
soil
Science society
potential.
Geoderma

34,
101-1
14.
Research
Organisation
2, 797-802.
ofAmerica Journal
50,841-848.
HETTIARATCHI, D.R.P.
&
FERGUSON, C.A. 1973.
Stress deformation behaviour of soil in root growth
Research
IS, 309-320.
in granular media.
Australian Journal
of
Soil
HETTIARATCHI, D.R.P.
&
~'CALLAGHAN, J.R. 1974.
Research
249
393-404.
A membrane model of plant cell extension.
Journal
ROMKENS, M.J.M.
~4
MILLER, R.D. 1971. Predicting
of

Theoretical Biology
45,459-465.
root size and frequency from one-dimensional con-
HETT~ARATCH~,
D,R,P,
&
OcALLAGHAN,
J,R,
1978, solidation data-a mathematical model.
Plant
and
Structural mechanics of plant cells.
Journal
of
Theoretical Biology
14,235-257. RUSSELL, R.S., 1977.
Plant
Root
Systems: Their
HOGETSU, T. 1986. Orientation
of
wall microfibril
Function and Interaction wifh rhe Soil.
McGraw-
deposition in root cells of
Pisum sativum
L.
var. Hill, London.
Alaska.
Plant Cell Physiology

27,947-951, RUSSELL, R.S.
&
Goss,
M.J. 1974. Physical aspects of
JAMIESON,
J.E., MORRIS, R.J.
&
MULLINS, C.E. 1988. soil fertility-the response of roots to mechanical
Effect of subsoiling
on
physical properties and crop impedance.
Netherlands Journal
of
Agricultural
growth on a sandy soil with a naturally compact
Science
2% 305-318.
subsoil.
Proceedings
of
the Zlth Conference
of
the
SCHOLEFIELD, D.
&
HALL, D.M. 1985. Constricted
International
Soil
Tillage Research Organisation
2, growth of grass roots through rigid pores.

Plant and
499-503. Soil85,153-162.
KAYS,
S.J.,
NICKLOW, C.W.
&
SIMONS, D.H. 1974. SMITH, K.A.
&
ROBERTSON P.D. 1971. Effect
of
Ethylene in relation to the response of roots to ethylene on root extension in cereals.
Nature
234,
LINDBERG,
S.
&
PETERSON,
S.
1985. Effects ofmechan- SOUTY, N. 1987. Aspect mkhanique de la croissance
ical stress
on uptake and distribution of nutrient in des racines.
I.
Mesure de la force de ptnetration.
barley.
Plant
and
Soil
83,295-309.
Agronomie
7,623-630.

REED, R.H. 1984.
Useandabuseofosmo-terminology.
Celland
Enviromenr7~
165-170.
B.G.
8~
GREACENy
E.L. 1986.
mechanics,
Journal
of
Agricultural Engineering
stresses on an expanding root analogue
237-248.
physical impedance.
Plant and
Soil40,565-571.
148-149.
358
A.
G.
Bengough
&
C.
E.
Mullins
(Received24
May
1989; accepted29 March

1990)
STOLZY, L.H.
&
BARLEY, K.P. 1968.
Mechanical
resistance encountered by roots entering compact
soils.
Soil Science 105,297-301.
TAYLOR,
H.M.
&
GARDNER, H.R. 1963.
Penetration
of
cotton seedling tap roots as influenced by bulk
density, moisture content and strength of soil.
Soil
Science 96, 153-156.
TAYLOR,
H.M.
&
RATLIFF, L.F. 1969.
Root elongation
rates
of
cotton and peanuts as a function
of
soil
strength and water content.
Soil Science 108,

113-119.
TOLLNER, E.W.
&
VERMA, B.P. 1984.
Modified cone
penetrometer
for
measuring soil mechanical impe-
dance.
Transactions
of
the American Society
of
Agricultural Engineers 27,33 1-336.
VEEN, B.W. 1982.
The influence of mechanical impe-
dance
on the growth of maize roots.
Rant
and Soil
66,101-109.
VESIC,
A.S. 1972.
Expansion of cavities in infinite soil
mass.
Journal
of
Soil Mechanics and Foundations
Division, American Society
of

Civil Engineers
VOORHEES, W.B., FARRELL, D.A.
&
LARSON,
W.E.
1975.
Soil strength and aeration effects
of
root
98(SM3), 265-290.
elongation.
Soil Science Society
of
America Pro-
ceedings 39,948-953.
WALDRON, L.J.
&
CONSTANTIN, G.K. 1970.
Soil resist-
ance to a slowly moving penetrometer.
Soil Science
WHITELEY, G.M.
&
DEXTER,
A.R. 1981.
The depen-
dence ofsoil penetrometer pressure
on
penetrometer
size.

Journal
of
Agricultural Engineering Research
WHITELEY,
G.M.,
UTOMO, W.H.
&
DEXTER,
A.R.
198
I.
A
comparison
of
penetrometer pressures and
the pressures exerted by
roots.
Plant and Soil 61,
351-364.
WHITELEY, G.M. &DEXTER, A.R. 1983.
Behaviour
of
roots in cracks between
soil
peds.
Plant and Soil74,
WILSON, A.J., ROBARDS, A.W.
&
Goss,
M.J.

1977.
Effects
of
mechanical impedance on
root
growth in
barley,
Hordeum vulgare
L.
11.
Effects on cell devel-
opment in seminal roots.
Journal
of
Experimental
Botany 28,1216-1227.
YAPPA, L.G.G., FRITTON,
D.D.
&
WILLATT, S.T.
1988.
Effect
of
soil strength
on
root growth under
different water conditions.
Plant and Soil 109,9-16.
109,221-226.
26,467-476.

153-162.