Plant Availability of Water in Soils Being Reclaimed from the ...

Plant Availability of Water in Soils Being Reclaimed from the Saline-Sodic State

Thesis submitted by

Nguyen Duy Nang

BSc (Agronomy), University of Agriculture and Forestry, Vietnam MSc (Soil Science), University of the Philippines, Philippines

For the degree of Doctor of Philosophy

in the

School of Agriculture, Food and Wine Faculty of Sciences

The University of Adelaide Adelaide, South Australia, Australia

December 2012

ii

Table of contents Page

List of tables ....................................................................................................................................... v

List of figures .................................................................................................................................... vii

Abstract ............................................................................................................................................ xii

Declaration ........................................................................................................................................xv

Acknowledgment .............................................................................................................................. xvi

Chapter 1 Introduction and literature review ............................................................................... 1

1.1 Introduction .......................................................................................................................... 1

1.2 Literature review .................................................................................................................. 2

1.2.1 Factors affecting plant available water in soils.................................................................. 2

1.2.2 Factors affecting soil structure .......................................................................................... 9

1.2.2.1 Exchangeable cations ................................................................................................. 9

1.2.2.2 Organic matter ......................................................................................................... 12

1.2.2.3 Clay swelling and dispersion .................................................................................... 13

1.2.2.4 Nutritional effects ..................................................................................................... 15

1.2.3 Effects of soil structure on soil strength, aeration, and hydraulic conductivity ............... 16

1.2.4 Salt-affected soils and their reclamation ......................................................................... 18

1.2.5 Models of plant available water ....................................................................................... 21

1.2.5.1 Historical models ...................................................................................................... 21

1.2.5.2 Integral Water Capacity (IWC) model ..................................................................... 25

1.2.5.3 Example of IWC calculations ................................................................................... 26

1.2.6 Conclusions ..................................................................................................................... 35

1.3 Overall problem, research questions and hypotheses. ........................................................ 36

1.3.1 Research questions .......................................................................................................... 36

1.3.2 Hypotheses ...................................................................................................................... 37

Chapter 2 Variation in soil water availability down the profile of a saline soil

using the Integral Water Capacity (IWC) model ........................................................................ 38

2.1 Introduction ........................................................................................................................ 38

2.2 Materials and Methods ....................................................................................................... 39

2.2.1 Site selection and sample collection ................................................................................ 39

2.2.2 Saturated hydraulic conductivity, water retention, and soil penetration resistance. ........ 42

2.2.3 Salinity and osmotic stress .............................................................................................. 43

2.3 Result and discussion ......................................................................................................... 44

2.3.1 Saturated hydraulic conductivity ..................................................................................... 44

2.3.2 Water retention curves ..................................................................................................... 44

iii

2.3.3 Soil penetration resistance............................................................................................... 46

2.3.4 Salinity and osmotic stress .............................................................................................. 47

2.3.5 Weighting functions ........................................................................................................ 49

2.3.5.1 Weighting the differential water capacities for salinity ............................................ 49

2.3.5.2 Weighting the differential water capacity for high soil penetration resistance. ....... 50

2.3.5.3 Weighting the differential water capacity for poor soil aeration. ............................. 52

2.3.5.4 Weighting the differential water capacity for declining soil hydraulic

conductivity. .......................................................................................................................... 54

2.3.6 Summarizing the effects of weighting the water capacity .............................................. 56

2.4 Conclusions ........................................................................................................................ 60

Chapter 3 In situ response of plants to saline conditions in the field ......................................... 61

3.1 Introduction ........................................................................................................................ 61

3.2 Materials and methods ....................................................................................................... 62

3.2.1 Experimental design ........................................................................................................ 62

3.2.2 Water balance model ....................................................................................................... 64

3.2.3 Calibrating the CPN 503 Hydroprobe neutron moisture meter. ...................................... 68

3.3 Results and discussion ....................................................................................................... 71

3.3.1 Plant water use from full canopy establishment to plant death ....................................... 71

3.3.2 Plant water use and root distribution ............................................................................... 72

3.3.3 Evaluation of the IWC model against water use by real plants....................................... 76

3.4 Conclusions ........................................................................................................................ 80

Chapter 4. Changes in IWC during reclamation of a salt-affected soil ..................................... 81

4.1 Introduction ........................................................................................................................ 81

4.2 Materials and Methods ....................................................................................................... 82

4.2.1 Experimental approach and design ................................................................................. 82

4.2.2 Experimental units .......................................................................................................... 83

4.2.3 Experimental protocol ..................................................................................................... 84

4.3 Result and discussion ......................................................................................................... 86

4.3.1 Changes in saturated hydraulic conductivity during reclamation ................................... 86

4.3.2 Changes in water retention curves during reclamation ................................................... 88

4.3.3 Changes in soil penetration resistance during reclamation ............................................. 93

4.3.4 Changes in IWC during reclamation ............................................................................... 96

4.4 Conclusions ...................................................................................................................... 101

Chapter 5 Shape of the salinity weighting function, ��o(h), based upon early plant

response to osmotic and matric stresses ...................................................................................... 102

5.1 Introduction ...................................................................................................................... 102

iv

5.2 Materials and Methods ..................................................................................................... 103

5.3 Result and discussion ....................................................................................................... 106

5.3.1 Dry matter yield as a function of osmotic-, matric- and total water potential ............... 106

5.3.2 A plant-based weighting function to attenuate the water capacity ................................ 107

5.4 Conclusions ...................................................................................................................... 113

Chapter 6 General discussion and directions for future research ........................................... 115

6.1 Introduction ...................................................................................................................... 115

6.2 Major findings (and future research) ................................................................................ 116

Appendices .................................................................................................................................... 121

References ..................................................................................................................................... 141

v

List of tables Table 1.1 Available water capacity by soil texture ............................................................................ 3

Table 1.2 Criteria used to classify salt-affected soils in U.S. and Australia. .................................... 19

Table 1.3 Summary of physical restrictions on the differential water capacity and their

effect on IWC.......................................................................................................................... 35

Table 2.1 Fitting parameters, k0, k1 and n, for the water retention curves in each horizon

down the soil profile. Optimization of the fitting parameters was conducted with

fixed (measured) values of the volumetric water content at saturation, �s, and

permanent wilting point, �wp. .................................................................................................. 46

Table 2.2 Fitting parameters for Equation [2.1] describing the relation between soil

penetration resistance (MPa) and soil matric head (cm), plus the matric heads, hi

and hf, respectively, at which SR(hm) reached values of 0.5 and 2.5 MPa. ............................ 47

Table 2.3. Measured values of the electrical conductivity of 1:5 soil:water extracts and

gravimetric water contents at saturation, plus the corresponding electrical

conductivity of paste extracts (calculated from Slavich and Petterson (1993)) and

values of hos and hm at wilting point (calculated from Equation 12 in Groenevelt et

al. (2004)). .............................................................................................................................. 49

Table 2.4. Predictions of plant available water in a saline soil profile based upon various

degrees of weighting of the differential water capacity; integrals at the top of each

column indicate the type of weighting applied: PAW = classical approach with no

attenuation, IWC = integral water capacity with attenuations to account for,

respectively: salt alone, salt + poor aeration, salt + high soil resistance, salt +

declining hydraulic conductivity, and all factors combined. .................................................. 58

Table 3.1 Chemical properties (saturated paste extracts) of the soil profile in the plots

containing the neutron access tubes. ....................................................................................... 69

Table 3.2 Variation in the standard 15 second slow neutron count rate with salt

concentration. .......................................................................................................................... 70

Table 3.3. Correlations between relative slow neutron count rate, RCR, and volumetric

water content, �, at each depth in the soil profile. .................................................................. 72

Table 3.4. Predictions of plant available water in a saline soil profile (mm/m) based upon

various degrees of weighting of the differential water capacity (taken directly from

Table 2.3) compared with field-measured change in water contents with Rhodes

grass (Chloris gayana cv. Pioneer). ........................................................................................ 76

Table 3.5. Predictions of plant available water (mm/m) based upon the same weightings

of the differential water capacity but ignoring salt, compared with field-measured

change in water contents with Rhodes grass (Chloris gayana cv. Pioneer) ........................... 77

vi

Table 4.1. Bulk densities achieved for 150 mm columns of soil from each soil horizon

(cylindrical pot diameter = 152.5 mm; calculated volume of each soil column =

2740 cm3). .............................................................................................................................. 84

Table 4.2 Elemental analysis by ICP-MS for the major cations, plus SAR, EC and pH of

the saturation paste extracts in each of the 9 soil horizons. SAR was calculated by

dividing [Na] (mmol/L) by the square root of ([Ca] + [Mg]). The value for �

cations (mmolc L-1) was the sum of ([Na] + [K]) plus twice the sum of ([Ca] +

[Mg]). The values of ECmeas were measured and they compare well with the values

for ECcalc, which were calculated from � cations divided by 10. ........................................... 84

Table 4.3 Fitting parameters for the Groenevelt et al. (2001; 2004) soil water retention

curves shown in Figures 4.4, plus the penetration resistance curves for the 6

different leaching treatments in the 9 soil horizons shown in Figures 4.5. ............................ 92

Table 4.4. Integral water capacity, IWC, of each soil horizon after leaching with solutions

of differing salinity and sodicity. IWC calculated using different weighting

functions. Initial EC for each horizon indicated in parentheses (dS m-1). Shadings

in 3rd & 4th columns indicate treatments where data were combined to form single,

average water retention or soil resistance curves. .................................................................. 97

Table 5.1 Numbered list of osmotic and matric potentials/heads used in soil and solution

culture. Osmotic potentials were calculated from the ECe values. ....................................... 104

Table 5.2 Parameters and constants for Equation [5.1] to describe the relative growth,

G′(hom), of Faba beans and Rhodes grass. ............................................................................ 109

Table 5.3. Parameters from Equation [5.2] used in preparing a weighting function to

attenuate the water capacity for salinity, based upon soil and plant factors

combined. In this study, the initial onset of osmotic stress was deduced to occur

from hi = 0.0025 bar for all examples. The value of hf in this table is the matric

potential at which wilting occurs under the salinity conditions corresponding to the

ECe; values were calculated from Equation (12) of Groenevelt et al. (2004). Colour

shaded data are shown in Figure 5.4 above. ......................................................................... 111

Table 5.4 Estimates of plant available water in soil of varying salinity based on soil

properties, or a combination of soil properties and plant response for Faba beans

and Rhodes grass. ................................................................................................................. 113

vii

List of figures

Figure 1.1 Relative hydraulic conductivity as a function of matric head for coarse-

textured and fine-textured soils. ................................................................................................ 4

Figure 1.2a “Divisions for classifying crop tolerance to salinity” (after Maas and

Hoffman 1977). ......................................................................................................................... 7

Figure 1.2b Response of some grain crops (e.g. rice, corn, wheat, barley) to salinity (after

Maas and Hoffman 1977). ........................................................................................................ 7

Figure 1.3 Change in aggregate tensile strength with aggregate diameter as a function of

time (after Lal and Shukla 2004) ............................................................................................ 17

Figure 1.4 The non-limiting water range (NLWR) of water contents as influenced by

restricting soil factors for plant growth in soil with (a) good structure and (b) poor

structure (Letey 1985). ............................................................................................................ 24

Figure 1.5 Effect of increasing bulk density on the water content at which volumetric air

content = 0.10m3/m3 and soil resistance = 2 MPa, superimposed on the water

contents at FC and PWP (after da Silva et al.(1994)); shaded area represents

LLWR. .................................................................................................................................... 24

Figure 1.6 Differential water capacity, C(hm), for the wet end (—); effective differential

water capacity, EK(hm), when a hydraulic conductivity weighting function ,ωK(hm),

is applied (– –); effective differential water capacity, EKa(hm), when both ωK(hm)

and a aeration weighting function, ωa(hm), are applied(Groenevelt et al. 2001)..................... 25

Figure 1.7 Weighted differential water capacities for a loamy soil accounting for salt-free

conditions and for conditions where EC of the saturated soil = 1, 2, 4, and 7.2 dS/m

(Groenevelt et al. 2004). ......................................................................................................... 26

Figure 1.8 Representation of the water retention curve using data and model for a loamy

sand published in Groenevelt et al. (2004). ............................................................................ 28

Figure 1.9 Differential water capacities for the loamy sand of Figure 1.8 when the soil is

salt-free (solid blue line) and when the soil has an osmotic head of 2 m in its

saturated state (dashed red line). The dotted ellipse identifies the section of the

curves discussed in Figure 1.10. ............................................................................................. 31

Figure 1.10 Differential water capacities for salt-free soil (solid blue line, 1), Saline soil

with hos = 2 m (solid red line, 2), Saline soil with poor drainage (dashed red line

segment, 3), and Saline soil with poor drainage and high strength (dashed purple

line segment, 4). ..................................................................................................................... 32

Figure 1.11 Weighting function to attenuate the water capacity for the effect of poor soil

aeration between the matric heads of hm = 0.51 to 1.41 m. .................................................... 33

viii

Figure 1.12 Weighting function to attenuate the water capacity for the effect of

increasingly high soil penetration resistance between the matric heads of hm1 = 2 m,

hm2 = 5 m. ............................................................................................................................... 34

Figure 2.1 Roseworthy paddock C1 ................................................................................................ 40

Figure 2.2 Exposed soil profile. ....................................................................................................... 40

Figure 2.3 Collecting undisturbed soil cores down the profile in paddock C1 ................................ 40

Figure 2.4 Laboratory set-up to measure saturated hydraulic conductivity on undisturbed

soil cores prior to measuring their water retention curves using field-isotonic

solutions. ................................................................................................................................ 42

Figure 2.5 Saturated hydraulic conductivities of undisturbed soil cores down the soil

profile using isotonic solutions applicable to each depth (horizontal red bars are

standard errors. ....................................................................................................................... 44

Figure 2.6 Water retention curves for the 9 soil horizons examined in this study: a) 0 to

25 cm, b) 25 to 75 cm, c) 75 to 115 cm, and d) 115 to 150 cm. ............................................. 45

Figure 2.7 Soil penetration resistance (SR, MPa) as a function of matric head (hm, cm)

for the same soils presented in Figures 2.6. The data falling between the horizontal

green and red dashed lines represent conditions that increasingly restrict root

growth in the soil. ................................................................................................................... 48

Figure 2.8 Differential water capacities for the nine water retention curves shown in

Figure 2.1 weighted (dotted lines) or not weighted (solid lines) for salt content

according to Groenevelt et al. (2004). .................................................................................... 51

Figure 2.9 Three possible shapes for weighting functions to attenuate the water capacity

based upon the ability of different plants to exert higher or lower root growth

pressures on their surroundings. Upper dotted lines come from using � = 0.2 (for

strong plant roots), solid lines come from using � = 0.5 (for medium-strength plant

roots), and lower dash-dotted lines come from using � = 1 (for weak plant roots). .............. 52

Figure 2.10 Three possible shapes (of many) for weighting functions to attenuate the

water capacity for poor soil aeration by varying the A-parameter in Equation [2.5]

from 0.2 (upper dotted lines), 0.5 (middle solid lines), and 1.0 (lower dash-dot

lines) according to the ability of different plants to tolerate poor soil aeration...................... 54

Figure 2.11 Three possible shapes for weighting functions to attenuation the water

capacity based upon the ability of different plants to cope with declining hydraulic

conductivity in dry soils. The lowest dotted lines come from using � = 0.2 in

Equation [2.9] for sensitive plants; the highest dash-dot lines come from using � =

1.0 for tolerant plants, and the central solid lines come from using � = 0.5 for

medium plants. ....................................................................................................................... 57

ix

Figure 2.12 Amount of plant available water down the profile of a saline soil (mm/m)

predicted by taking into account different soil physical restrictions listed in Table

2.5. .......................................................................................................................................... 59

Figure 3.1 Diagram of experimental plots showing dimensions and locations of neutron

access tubes. ............................................................................................................................ 62

Figure 3.2 Preparation of the three isolated field plots for complete profile saturation and

planting of Rhodes grass (Chloris gayana cv. Pioneer). ......................................................... 63

Figure 3.3 Water supply system, rain-shelter frame, taking readings with neutron probe. .............. 65

Figure 3.4 Canvas suspended from rain shelter to shed any rain when expected (not

often). ...................................................................................................................................... 66

Figure 3.5 Photographs of the perennial Rhodes grass (Chloris gayana cv. Pioneer) plots

from the last irrigation (27 Jan 2011) until the plants stopped extracting water and

never recovered after rainfall (15 June 2011). ........................................................................ 67

Figure 3.6 Mean standard 15 second count rate, CRs, of CPN 503 Hydroprobe in a large

drum of water having different salt concentrations as measured by EC (dS m-1).

The red vertical bars through each point represent the ± standard error of the mean

of 20 readings.......................................................................................................................... 70

Figure 3.7a Volumetric water content as a function of depth for plots 1 and 2, from the

time of the initial profile saturation (03 Nov 2010) until the plants wilted

completely (15 June 2011). Horizontal bars represent ± standard error of the mean

water content. .......................................................................................................................... 73

Figure 3.7b Volumetric water content as a function of depth for plots 3 and 2, from the

time of the initial profile saturation (03 Nov 2010) until the plants wilted

completely (15 June 2011). Horizontal bars represent ± standard error of the mean

water content. .......................................................................................................................... 74

Figure 3.8 Distribution of Rhodes grass root mass per unit volume as a function of depth

below the soil surface. ............................................................................................................ 75

Figure 3.9. Real soil water extraction for Rhodes grass (Chloris gayana cv. Pioneer)

superimposed on estimates of water availability after various weightings of the soil

solution capacity (i.e. including consideration of salt). .......................................................... 77

Figure 3.11. Comparative estimates of water availability from Figure 3.10 adjusted with

‘gentler’ coefficients in the weighting functions. ................................................................... 79

Figure 3.12. Comparative estimates of water availability from Figure 3.11 adjusted with

significantly ‘gentler’ coefficients in the weighting functions. .............................................. 79

Figure 4.1 Dimensions of experimental pot of soil with 4 small soil cores embedded. ................... 83

x

Figure 4.2 Leaching and sampling protocol for each soil horizon. Treatment numbers are

indicated in the first pot on the left......................................................................................... 85

Figure 4.3 Changes in saturated hydraulic conductivity of repacked soil from a profile

using leaching solutions of different EC and SAR. ................................................................ 87

Figure 4.4 Summary of water retention curves grouped according to whether treatment

effects were obvious for soil horizons: a) 0-10 cm, b) 10-25 cm, c) 25-35 cm, and

d) 35-55 cm. Groupings of curves are indicated for each soil horizon. ................................. 89

Figure 4.4 Water retention curves grouped according to whether treatment effects were

obvious for soil horizons: e) 55-75 cm, f) 75-100 cm, g) 100-115 cm, and h) 115-

150 cm. Groupings of curves are indicated for each soil horizon. ......................................... 90

Figure 4.4 Water retention curves grouped according to whether treatment effects were

obvious for soil horizon: i) > 150 cm. Groupings of curves are indicated. ............................ 91

Figure 4.5 Soil resistance curves grouped according to whether treatment effects were

obvious for soil horizon: a) 0-10 cm, b) 10-25 cm, c) 25-35 cm, and d) 35-55 cm.

Groupings of curves are indicated for each soil horizon. ....................................................... 94


obvious for soil horizon: e) 55-75 cm, f) 75-100 cm, g) 100-115 cm, and h) 115-

135 cm. Groupings of curves are indicated for each soil horizon. ......................................... 95


obvious for soil horizon: i) > 150 cm. Groupings of curves are indicated. ............................ 96

Figure 4.6 Profiles of soil water availability calculated by weighting the water capacity

of the soil in its initial saline state for different limiting factors. ........................................... 98

Figure 4.7 Increases in plant available water (IWC) during reclamation of the soil profile

from its initially sodic-saline state to a calcic non-saline state calculated using a)

only the osmotic weighting function of Groenevelt et al. (2004), and b) all

weighting functions. NB. The scales on the available water axis for parts a) and b)

are different. ......................................................................................................................... 100

Figure 5.1 Soil water retention curve of Monarto soil packed at a bulk density of 1.4 g

cm-3. Parameter values for the Groenevelt et al. (2004) equation are: �s = 0.405, �wp

= 0.100; k0 = 0.409; k1 = 0.328, and n = 0.646. ................................................................... 105

Figure 5.2 Mean whole-plant dry matter yield per plant plotted as a function of the total

hydraulic potential (absolute value) for Faba beans and Rhodes grass grown in

solution culture and soil. The vertical bars represent standard deviations of each

mean point. ........................................................................................................................... 106

xi

Figure 5.3 Relative growth, G′(hom), as a function of the total hydraulic potential of the

water for a) Faba beans, and b) Rhodes grass. The red-dashed line represents

Equation [5.1], the parameters for which are given in Table 5.1. ......................................... 108

Figure 5.4 Four examples of weighting functions (Equation [5.2]) to account for salt in

the soil water for a) Faba Beans, and b) Rhodes Grass. The Blue, Green, Brown

and Red lines are for Pots 1, 11, 23 and 37 respectively (colour-coordinated data

highlighted in Table 5.3). ...................................................................................................... 110

Figure 5.5 Differential water capacity (solid black line) with 4 x examples of effective

water capacities superimposed for Pot 1 (solid blue line), Pot 11 (dashed green

line), Pot 23 (dash-dot brown line), and Pot 37 (solid red line) for a) Faba beans

and b) Rhodes grass. ............................................................................................................. 112

xii

ABSTRACT

The work reported in this thesis was motivated by a desire to improve our ability to estimate the

amount of plant available water in soils beyond the classical methods enveloped in the terms “Plant

Available Water” (PAW) and “Least Limiting Water Range” (LLWR). It took the view that soils

can be considered to be water ‘capacitors’ that are influenced primarily by the physical properties

of the soil. The soil properties of particular concern in this work were the soluble salt concentration

in the soil water, poor soil aeration in wet soils, rising penetration resistance and declining

hydraulic conductivity in drying soils. Their effects on soil water availability were embodied the

model proposed by Groenevelt et al. (2001; 2004) called the integral water capacity (IWC). The

theoretical framework for the IWC-model is quite strong, if not intuitive, but there is little

published evidence to date to evaluate its integrity using real plants growing in real soils. There is

also little information to enable one to calculate plant available water in soils being reclaimed from

the saline-sodic state. The work reported in this thesis therefore aimed to address four main

questions:

Question 1 (Chapter 2)

When soil salinity, aeration, strength and hydraulic conductivity are all taken into account, how

much soil water is available to nominally ‘salt-sensitive’ plants when calculated using the IWC

model of Groenevelt et al. (2004)?

Undisturbed soil samples were collected from the profile of a saline soil whose texture gradually

became heavier with depth. Water retention, soil resistance, soil aeration and soil salinity were all

measured and used to prepare appropriate weighting functions to attenuate the differential water

capacity and obtain different estimates of plant available water down the soil profile. All weighting

functions attenuated the water capacity and reduced the IWC to varying degrees, each of which

produced smaller estimates of plant available water than the classical PAW model. Weighting due

to salinity caused by far the greatest individual reduction in IWC, followed by soil resistance, soil

aeration, then hydraulic conductivity. The combination of all factors, of course, reduced IWC the

most. However, replication of the findings (and therefore a statistical evaluation of the effects) was

not possible, so these findings must be treated as tentative for now. Furthermore, many of the

weighting functions were applied with little or no knowledge of the real magnitude of their

parameters based upon real plant behaviour. To take this into account, weighting functions were

proposed for each limiting soil property having functional forms that included plant-specific

parameters, whose magnitude can be varied widely for different plants. The plant-specific

xiii

parameters attenuate the water capacity severely when a plant species is sensitive to a restricting

soil property and not as severely when a plant species is not sensitive to it.


To what extent do the (laboratory-based) estimates of soil water availability using IWC coincide

with the (field-based) measurements of soil water use by real plants?

A field experiment was conducted on a saline soil, in which a water budget was constructed to a

depth of 1.5 m and a crop of relatively salt-tolerant perennial Rhodes grass (Chloris gayana cv.

Pioneer) was grown to full canopy before stopping all water inputs. The volumetric water content

was monitored regularly (using a specially calibrated neutron moisture meter) as the crop

transpired water over several months until it eventually died from water stress. The total change in

water content down the profile was determined by the difference in water contents at the time of

saturation and those at the time of permanent plant wilting. The predicted and measured amounts of

available water were compared with the classical PAW model and it was concluded that the

magnitude of attenuation proposed by Groenevelt et al. (2004) was too severe. Some effort was

made to adjust the plant-specific slope parameters, �, A, and �, but with no real knowledge of the

magnitude of these parameters for different plants, it was considered futile to expend much time

adjusting the parameters without new information about real plants.


When saline-sodic soils are ‘reclaimed’ toward the non-saline, calcic state, to what extent does soil

water availability change (in terms of IWC) without significant soil disturbance in the process?

A column-leaching experiment was conducted in the laboratory using re-packed soil cores leached

first with a saline solution (isotonic with field conditions) then with various different salt solutions

to determine the extent to which changes in the pore size distribution would be accompanied by

measurable changes in salinity, soil strength, hydraulic conductivity and aeration – and thus, plant

available water (IWC). Fifty-four different average water retention curves were prepared in this

experiment, and the curves were differentiated to produce water capacities that were weighted

according to procedures outlined in Chapter 2. As in Chapter 3, it was found that the salinity

weighting function caused the greatest reduction in IWC and was probably too severe. It was also

found that the other factors reduced the water capacity somewhat, in declining order of importance:

salt > aeration > strength > hydraulic conductivity. It was a surprise to find that with no disturbance

xiv

of the re-packed soil samples, the structure of the soil was able to be changed to a small extent

without disturbing it mechanically, simply by changing the composition of the leaching solution.


To what extent does the response of plants to increasing salt concentration mimic the peculiar

shape of the weighting function proposed by Groenevelt et al. (2004)?

Plants of two different types (Faba beans, Vicia faba cv. Fiord; and Rhodes grass, Chloris gayana

cv. Pioneer) were grown in a glasshouse in either pots of salt-solutions or in soil having different

salt concentrations. The idea was to develop a weighting function for salinity based upon measured

plant growth responses to varying salinity, and compare this with the peculiarly shaped weighting

function for salt proposed by Groenevelt et al. (2004). The growth reduction pattern due to salt was

similar for both plants, so the relative growth of each plant was plotted as a function of the total

water potential. It was found that the relative growth of the solution-grown plants coincided with

those for the soil-grown plants, which implied the plants responded in the same way to both

osmotic and matric stresses. Relative growth responses were then fitted to a (rather inadequate)

model, which was then used in a weighting function that included both plant- and soil-specific

fitting parameters. The results produced a more gentle attenuation of the water capacity than the

model of Groenevelt et al. (2004), which suggests there is considerable room to adjust the

‘reflection coefficient’ in their model. Finally, the typical ‘bent-stick’ model used to describe plant

response to salinity was found to be out-dated and should be replaced by a more modest, smooth

decline in plant growth with increasing salt concentration.

xv

Declaration

This work contains no material which has been accepted for the award of any other degree

or diploma in any university or other tertiary institution and, to the best of my knowledge

and belief, contains no material previously published or written by another person, except

where due reference has been made in the text.

I give consent to this copy of my thesis, when deposited in the University Library, being

made available for loam and photocopying, subject to the provisions of the Copyright Act

1968.

I also give permission for the digital version of my thesis to be made available on the web,

via the University’s digital research repository, the Library catalogue, the Australian

Digital Thesis Program (ADTP) and also through web search engines, unless permission

has been granted by the University to restrict access for a period of time.

Dated:___________________ ______________________________

Duy Nang Nguyen

xvi

Acknowledgment

I would like to sincerely thank my supervisors Dr Cameron Grant and Dr Robert Murray for their

advice, sharing knowledge, valuable discussions and constructive criticism. I am very grateful to

both of them for providing me with the necessary facilities to carry on my research and during the

preparation of my thesis. I specially thank my principal supervisor Dr Cameron Grant who has

brought me throughout states of PhD life with full support, taking care and understanding my

difficult time and finance, without his support and encouragement I would never have reached this

state.

I would also like to sincerely thank my close friend Jennifer McKeon for her concern, support and

encouragement throughout four years of my study. With all my heart I have to say without her

support for taking care of my son I would have had much less time to study.

I would like to have a big thank go to Leonie McKeon and Shelley Rogers for all their concern,

support and encouragement throughout the years of my study.

I would like to thank Keith Cowley, Country Fire Service – Mudla Wirra, for valuable help with

supplying water to irrigate my field plots for eight months of experimental work.

Special thanks go to Hugh Cameron (Workshop co-ordinator at Roseworthy campus) who allowed

me to use many pieces of equipment during my field work at Roseworthy campus.

I would like to thank the Ministry of Education and Training of Vietnam for providing me a

scholarship which enabled me to undertake and complete this study. Also, special thanks to The

University of Adelaide for waving my tuition fee for the first semester 2012 which allowed me to

finish writing up my thesis.

I am deeply indebted to my parents for their love, long-term support and encouragement in all

aspects of my career and life.

Ultimately, I would like to thank my lovely son, Bao, firstly for his love, patience and

understanding during 4 years of my study, and secondly for taking care of himself during the time

his daddy frequently stayed back late to write the thesis.

1

Chapter 1 Introduction and literature review

1.1 Introduction

During the reclamation of saline land by leaching, excessive swelling and dispersion of soil

colloids can occur if soil electrolytes are not managed carefully. This invariably causes dramatic

structural decline leading to reduced drainage and aeration, increased resistance to root

penetration and thus reduced soil water availability to plants. The amount of water available to

plants will increase only after soil physical properties improve but a quantitative evaluation of

such increases has yet to be published. Calculating the amount of water that becomes available

to plants during reclamation is not simple because soil physical properties change and the

changes have complex effects on soil hydraulic properties.

The classical model to calculate plant available water (PAW) integrates the water content held

in the soil between field capacity (FC) and permanent wilting point (PWP) as shown in the

relation (Gardner 1960):

[1.1]

Of course, the PAW concept is too simplistic because it ignores all the effects of soil physical

and chemical properties affecting soil water extraction by plants. The model of Groenevelt et al.

(2001) calculates the amount of soil water potentially available to plants by applying weighting

functions to the differential water capacity and then integrating to produce an integral water

capacity, IWC. The IWC, however, remains rather theoretical and provides only a crude

estimate of water availability because it has never been checked against real plants in real soils

undergoing reclamation.

The presence of salt in soil is also a principal constraint that limits plant access to soil water,

because it decreases the osmotic potential and thus the total potential of water in the soil

solution. Different plant species have different ways to cope with osmotic effects of the soil

solution. Plants with high salt tolerance can overcome the osmotic stress by various mechanisms

so that their roots only face water stress as the matric potential drops. However, salt-sensitive

plants do not have special mechanisms to cope with osmotic stress so their growth rates in

saline environments reflect their ability to cope with osmotic stress.

Plants grow well in the presence of modest concentrations of soluble salts (e.g. nutrient salts

from fertilizers) but they begin to experience water stress as the salt concentration increases. An

2

analysis of data showing plant response to salinity (Maas and Hoffman 1977) suggests there is

an abrupt point at which plant growth declines linearly to zero with increasing salt concentration

(sometimes known as the ‘bent-stick’ model), although it is more likely that a gradual transition

occurs and that the end point is less rigid than originally thought (Steppuhn et al. 2005; Sheldon

2009). This study evaluates the integral water capacity (IWC) model using real plants grown

under saline conditions in the field and the laboratory. It also monitors changes in soil physical

properties and their effects on IWC during the process of reclamation. The observations of plant

response to different salinity will inform the choice of osmotic weighting functions to calculate

crop-specific estimates of soil water availability and thereby increase the general utility of the

IWC proposed by Groenevelt et al. (2001) in saline soil. The following Literature Review will

examine the historical development of the concept of soil water availability and the factors that

affect it, with special reference to what happens to water availability during reclamation of

saline sodic soils.

1.2 Literature review

1.2.1 Factors affecting plant available water in soils

The amount of soil water that is available to plants is largely controlled by the texture of the

soil, which dictates the general range of pore sizes. This is also controlled by the structural

arrangement of the particles, which is controlled, in turn, by many factors, including salinity and

exchangeable cations. It can also be argued that for plants to make use of the water held in soil

their roots need to be able to freely explore the soil, which depends on how hard the soil is and

how well aerated it is. In addition, the soil needs to be able to deliver water to the plant roots

upon demand, and this is controlled by the unsaturated hydraulic conductivity of the soil. The

role of each of these factors will be reviewed and then the potential to quantify their effects on

plant available water using the IWC will be explored.

Soil texture

Soil texture refers to the proportion by mass of sand-, silt-, and clay-sized particles and this

largely controls how much water can be stored in the soil. It is generally accepted that the water

holding capacity of coarse-textured (sandy) soils is much less than that of fine-textured (silty

and clayey) soils. This is because mixtures of large mineral particles produce soil matrices with

larger pores in them, which retain less water for plants to extract than soil matrices with smaller

particles. Large pores in sandy soils allow water to drain quickly under the influence of gravity

so this water is often lost from the root zone before plants can use it. The remaining water is

3

held in smaller pores by capillary forces, but sandy soils have a limited proportion of these

smaller pores so they do not store much water. By contrast, clayey soils contain many more tiny

particles which create large surface areas for water adsorption and a large volume of tiny pores.

These tiny pores hold water much more tightly, thus clay soils retain more water than do sandy

soils. However, because most plants can only exert up to approximately 1500 kPa of suction

(corresponding to pores > 0.2m diameter), a great deal of water is held in clay soils that plants

cannot extract. Table 1.1 shows the strong influence of textural class on the available water

capacity of the soil. The window of plant available water is narrow in sandy soils, widest in the

medium-textured soils and relatively narrower in the finer-textured soils. The proportion of soil

pores that occur in the range from which plants can extract water is controlled primarily by

texture, but is also influenced by the spatial arrangement of the particles, or the soil structure.

Hence although clay-textured soils can have relatively low available water capacity, this is only

a generalisation. The available water capacity of a clay soil can be greatly increased by

improving its structure.

Table 1.1 Available water capacity by soil texture

Source: Jeff (1997)

Another characteristic of coarse-textured soil that contributes to low soil water availability is

low unsaturated hydraulic conductivity. Saturated conditions rarely last very long in most

agricultural soils, and when this does occur, most of the water simply drains from the larger

pores under gravitational forces, after which the unsaturated hydraulic conductivity drops

precipitously (Figure 1.1). The unsaturated hydraulic conductivity is a nonlinear function of the

soil water content and the soil matric potential (or matric head when expressed in dimensions of

length). At lower matric heads (i.e. near saturation, or matric head 0) sandy soils have a

larger hydraulic conductivity than clayey soils. However, as these soils de-saturate, the

hydraulic conductivity in sandy soils decreases more readily than in the clayey soils because

sandy soils contain larger pores, which drain faster compared to clayey soils, which have

relatively smaller pores. Since a greater number of small pores in clayey soils are filled with

A NOTE:

This figure/table/image has been removed to comply with copyright regulations. It is included in the print copy of the thesis held by the University of Adelaide Library.

4

water the continuity of water-filled pores remains greater, the tortuosity of flow is smaller, and

so the hydraulic conductivity remains greater compared to that in sandy soils.

Figure 1.1 Relative hydraulic conductivity as a function of matric head for coarse-textured and fine-textured soils.

Soil structure

Soil structure can be defined in terms of form and stability (Kay 1990). Soil structural form

refers to the arrangement of soil particles into stable units called aggregates (Marshall et al.

1996), whereas the stability of soil structure is its ability to retain this arrangement when

exposed to different stresses (Angers and Carter 1996). The composition, size, and arrangement

of pore space located between aggregates are important factors contributing to water storage and

supply for plants. The ability of soil to transmit water depends on the presence of interlinked

pores and on their size and geometry. Pore diameters may range from < 0.2 μm to 10 mm or

more. Pores between 0.2 and 30 μm in diameter are important for storing soil water for later use

by plants. Pores < 30 μm allow rapid absorption of water into soil and pores between 30 and

300 μm are important for infiltration and drainage but generally do not retain water long enough

for plants to use it (Connolly 1998).

It has long been known (e.g.(Doneen and Henderson 1952; Quirk and Schofield 1955; Emerson

and Smith 1970; Bakker 1972) that soil structural stability depends mainly on four soil

properties: exchangeable cations (Ca, Mg, Na, K), electrolyte concentration (or salinity), pH,

and organic matter content. Conditions where there is a preponderance of monovalent

|Matric head|

Sandy soil

Clayey soil

Rel

ativ

e H

ydra

ulic

Con

duct

ivity

5

exchangeable cations (usually Na), low electrolyte concentration, high pH and low organic

matter content all lead to unstable soil structures in which aggregate slaking plus colloid

swelling and dispersion occur. By contrast, divalent exchangeable cations (especially Ca),

modest salt concentrations, neutral pH and high organic matter content all contribute to stable

soil structures. Stability of aggregates and the pores between them plays an important role in

movement and storage of water, aeration, biological activity and the growth of roots.

Aggregates that break down when wet have smaller pores, reduced pore continuity and

increased soil strength upon drying (Cresswell et al. 1992). For example, when aggregates

slake, swell and disperse, the average pore size decreases, which reduces infiltration and

hydraulic conductivity by as much as 2000 times (McIntyre 1958), increases surface crusting

and strength, which reduces root penetration and therefore effectively reduces the amount of

water available for plant extraction (Gupta et al. 1989; McGarry 1990; Lipiec et al. 1991).

Soluble salts

Plant available water may also be reduced by factors other than soil texture and structure,

namely osmotic effects and specific-ion effects (Groenevelt et al. 2004; Qadir et al. 2006).

Osmotic effects refer to elevated concentrations of soluble salt in the soil water which reduce the

water pressure gradient between the soil solution and plant root cells (Jensen 1982). The

movement of water into plant roots is a response to an osmotic pressure gradient induced by the

plants. So the closer the soil solution is to free water, the easier it is for root cells to draw in that

solution. In non-saline soils, osmotic effects are generally ignored in calculating soil water

availability. However, in saline soils, salt in the root zone makes the osmotic head of the soil

water, ho, greater in accordance with van’t Hoff’s law:

gcRTho �

,

[1.2]

where c is total concentration of dissolved species, mol m-3, R is universal gas constant, J K-1

mol-1, T is absolute temperature, K, � is density of water, kg m-3 and g is acceleration due to

gravity, m s-2. Therefore, although water may not always be tightly held by the soil matrix, the

presence of soluble salt forces plants to exert more energy to extract the water by overcoming

the osmotic head. Qadir et al (1996) reported that under saline/sodic conditions, osmotic effects

can reduce the amount of water entering roots even when the soil is at field capacity.

Rengasamy (2010) found that osmotic effects were responsible for reducing dry matter

production to 50% when the electrical conductivity of the soil solution exceeded 30 dS/m.

6

In addition to osmotic effects, specific ion toxicities sometimes occur in sodic and saline soils

due to the presence of excess concentrations of the cation, Na+, or the anion, Cl-, both of which

interfere with normal physiological cell function in plant roots (Robinson 1971; Zaitseva and

Sudnitsyn 2001). Borate is another ion that commonly occurs at toxic concentrations in solution

with sodium and chloride (Bell 1999). Martin and Koebner (1995) demonstrated that chloride

ion toxicities were partly responsible for the dramatic reduction in vegetative and reproductive

growth in Mexican bread wheat (cv. Glennson) when the plant was exposed to medium-to-high

concentrations of NaCl (180 mM).

Plants differ in their ability to survive and yield satisfactorily when grown in saline soils. There

is much literature on the relative tolerance of different crops to soil salinity obtained under a

vast range of soil, climatic and salinity conditions. Previous studies (Pearson 1959; Kaddah and

Fakhry 1961; Pearson 1961) have shown that tolerance to salinity is not a fixed property of a

species but something that varies with the growth stage of the plant, climatic conditions and

even within the same species for different varieties. Furthermore, the methods used by different

workers to study salt tolerance vary from water culture experiments to field studies where there

is little control over the root zone salinity. Maas and Hoffman (1977) compiled and reviewed

the available data describing the relative (not absolute) salt tolerance of different agricultural

crops (Figure 1.2a). These figures show that, in general, crop yields are not reduced

significantly until a threshold salinity is exceeded, and then the yields decrease approximately

linearly as the salinity continues to increase. The salt tolerance line for each crop was obtained

by calculating a linear regression equation for the yield beyond a threshold point, although the

way in which this threshold was established was rather arbitrary. It is more likely that a

polynomial would describe most crop-response functions (Steppuhn et al. 2005).

The data plotted in Figure 1.2a are for saline conditions only and do not necessarily encompass

the separate effects of sodicity. Barley, for example, is known to be tolerant of saline conditions

(Figure 1.2b) but not very tolerant of sodic conditions, which is mainly due to poor soil aeration

in sodic soils. Similarly cotton, while tolerant of saline conditions, is only moderately tolerant of

sodic conditions, or even sensitive to sodic conditions at some growth stages. On the other hand,

rice is only moderately sensitive to saline conditions (Figure 1.2b) but is highly tolerant of sodic

conditions (Pearson 1959; Kaddah and Fakhry 1961; Pearson 1961).

7

Figure 1.2a “Divisions for classifying crop tolerance to salinity” (after Maas and Hoffman 1977).

Figure 1.2b Response of some grain crops (e.g. rice, corn, wheat, barley) to salinity (after Maas and Hoffman 1977).

A NOTE:


A NOTE:


8

One strategy to improve productivity of salt-affected soils is to use a range of plants having

different tolerance to sodicity. Bernstein (1975) observed that the growth of sensitive species

may be affected by soils with an exchangeable sodium percentage (ESP) as low as 10, while

many crops are moderately tolerant of sodicity and may not be affected until the ESP rises to

25. Highly tolerant crops may not be affected until the ESP exceeds 50. Tolerance of high ESP

depends on the ability of the plant to take up Ca and Mg despite low concentrations in the soil

solution and significant competition by Na for uptake (Bernstein 1975; Kinraide et al. 2004).

To link the soil properties with the salt tolerance of different plants, Kopittke and Menzies

(2005) introduced the concept of calcium activity ratio, CAR. CAR is the activity of calcium in

the soil solution is divided by the sum of all cation activities in the soil solution:

, [1.3]

where parentheses denote activity of each cation in solution.

Critical CAR values associated with a 10% reduction in root length were found to be 0.025 for

Mungbean and 0.034 for Rhodes grass across a range of Na concentration and pH in both soil

and solution culture (Kopittke and Menzies 2005). As the soil dries, the concentration of salt in

the soil solution increases, further increasing the osmotic head. To maintain water uptake from a

saline soil, plants must osmo-regulate (Morgan 1984). This is done either by taking up salts and

compartmentalizing them within plant tissue, or by synthesizing organic solutes to generate a

competitive osmotic potential within the plant. Plants which take up salts generally have a

higher salt tolerance and greater ability to store high salt concentrations in plant tissue without

affecting cell processes and are known as halophytes. Plants which synthesise organic solutes

are known as glycophytes and they try to prevent excess salt uptake because they can only

tolerate much lower concentrations of salt in plant tissues before cell processes are adversely

affected. Increasing uptake of salts by halophytic plants to adjust osmotic potential may result in

Na+ and Cl- toxicities. Accumulation of excess Na+ may also cause metabolic disturbances in

processes where low Na+ and high K+ or Ca2+ are required for optimum function (Marschner

1995). A decrease in nitrate reductase activity, inhibition of photosystem II (Orcutt and Nilsen

2000), and chlorophyll breakdown (Krishnamurthy et al. 1987) are all associated with increased

Na+ concentrations. Cell membrane function may be compromised as a result of Na+ replacing

Ca2+, resulting in increased cell leakiness (Orcutt and Nilsen 2000).

9

A study of non-halophytic plant response to salinity (Munns and Termaat 1986) pointed out that

leaf growth is more sensitive than root growth. Regulating leaf expansion in this study was

probably caused by a message from the roots because of the water status (Munns and Termaat

1986). Further research into plant response to salinity indicated that there were two phases

involved in this response, as follows (Munns et al. 1995):

The first phase involved a large decrease in growth rate caused by the salt outside of the roots

(osmotic response) and the second phase was an additional decline in growth caused by ion

toxicity within the plant. Results of a trial that investigated the response of halophytic plants to

salinity showed that all genotypes had a similar reduction in growth with salt. After an initial

period, the more sensitive genotypes showed greater reduction in growth. These data strongly

support the hypothesis of a two-phase response and therefore indicate that any genotypic

differences in the first phase relate to the osmotic effect and not to an ion specific effect. The

second phase began only after toxic levels had accumulated in the leaves in sufficient quantities

to cause leaf necrosis and therefore a resultant reduction in available assimilate (Munns et al.

1995).

The second phase of saline irrigation resulted in a degree of yield reduction that was greatly

affected by the stage of plant development at the time when saline irrigation began. Plots that

received the high salt treatment at an early stage suffered the greatest level of damage,

presumably due to a greater ionic accumulation over time, while a significant interaction

between genotype and salt treatment affected grain yield in this second phase. Munns et al.

(1995) concluded that the length of the first phase was dependent on the concentration of salt in

the soil, transpiration rate and the ability of the genotype to exclude Na. Other researchers

(Delane et al. 1982; Cranner and Bowman 1991; Yeo et al. 1991; Neuman 1993) also found that

the initial response to salinity was dependent on water potential, rather than the specific salt.

1.2.2 Factors affecting soil structure

1.2.2.1 Exchangeable cations

The suite of exchangeable cations on the soil exchange sites has an enormous influence on soil

hydraulic properties. The primary four cations of interest are sodium, calcium, magnesium and

potassium. Each of these will be examined in turn.

10

Sodium

It is generally accepted that exchangeable sodium causes soils to be weakly aggregated, and the

amount of exchangeable sodium is frequently used to represent the physical condition of a soil

(Kemper and Koch 1966). The effect of exchangeable sodium differs depending on the part of

the profile in which it is found. For example, high amounts of exchangeable sodium in the

surface layers will generate poor physical characteristics (e.g. high strength and low

permeability) if this layer is subject to mechanical stress (Rowell et al. 1969).

As the exchangeable sodium content of soil increases, repulsive forces develop between

particles and there is an increasing tendency towards excessive swelling and dispersion of the

colloidal fraction and this destroys soil structure and blocks soil pores (Doneen and Henderson

1952; Baver 1956; Bakker 1972; Bronick and Lal 2005). Even in non-sodic soils, the quality of

soil structure may be diminished by soluble and exchangeable Na (Bakker 1972), depending on

soil pH, clay mineralogy and the presence of electrolytes in the soil solution. Sodic soils may

not disperse if there is sufficient salt to depress swelling and maintain flocculation (Baver

1956).

Many soils around the world exhibit adverse physical properties when ESP > 15 (Richards

1953), although for Australian soils ESP > 5 or 6 is sufficient to generate soil physical problems

(Northcote and Skene 1972; Mikhail 1974).

Calcium

It is widely recognized that exchangeable calcium contributes to a strongly aggregated soil; the

poor structural qualities of sodic soils can be significantly improved if exchangeable sodium is

replaced by both soluble and exchangeable calcium (Bakker 1972; Greene et al. 1978a;

Grierson 1978). Calcium and magnesium are both believed to exert positive effects on soil

structure via cationic bridging with clay and soil organic carbon (Bronick and Lal 2005).

However, Ca2+ is more effective than Mg2+ in improving soil structure (Zhang and Norton

2002).

Magnesium

The effect of exchangeable magnesium on soil aggregate stability is somewhat controversial.

On the one hand, Mg is divalent so it is generally more effective at reducing swelling and

dispersion on its own than say Na. The effects of Mg, however, are not driven by valence alone

11

– they also depend on the prevailing soil conditions. For example, wet, sheared aggregates from

surface soils of low pH were found by Emerson and Smith (1970) to be more susceptible to

dispersion if leached with MgCl2 as opposed to CaCl2; this was thought to be related to the

different effects of Mg and Ca on the solubility of soil organic matter. For example, in a

hydromorphic grey clay-textured soil, which contained mainly montmorillonite, dispersion

occurred when the Mg saturation was > 30% in the presence of organic matter (Bakker 1972).

Regardless of the organic matter content, Joffe and Zimmerman (1945) and Mikhail (1974)

found that soils relatively high in Mg behaved like soils with high exchangeable Na, while soils

with relatively more Ca than Mg could tolerate higher amounts of exchangeable Na without

dispersing. Zhang and Norton (2002) have also noted that Mg2+ may have a deleterious effect on

soil aggregate stability by increasing clay dispersion and that Mg2+ may result in increased

swelling by expanding clays, resulting in disruption of aggregates.

Also, the soil moisture content at which dispersion becomes evident depends on the nature of

the dominant exchangeable cations (Bakker 1972). For example, in remoulded Shepparton fine

sandy loam, Mg-saturated soil begins to disperse at a matric head close to 150 m, whereas Ca-

saturated soil begins to disperse at a significantly greater water content, corresponding to a

matric head of only 1 m. Also the water content for dispersion at a given ESP for a Ca-Na soil is

much greater than that for a Mg-Na soil. This illustrates that the attractive forces between clay

crystals for the Mg-Na system are very low even when the soil is relatively dry.

However, not all soils behave the same in respect to exchangeable cations. For example, Ahmed

et al. (1969) found no difference between the effects of Ca and Mg on aggregate stability and

hydraulic conductivity, while El-Swaify et al. (1970) found no differences in degree of

dispersion, liquid limit and moisture retention for a montmorillonitic soil. Similarly, Koenigs

and Brinkman (1964) considered that low stability originated mainly from a combination of

high exchangeable Na and low organic matter content rather than from high Mg.

Potassium

Potassium, K, can have a similar effect to Na on soil aggregation (Ahmed et al. 1969), and in

some soils the application of potassium fertilizer can lead to soil structural degradation. In such

cases, the addition of K fertilizer needs to be conducted only where there is plenty of organic

matter (e.g. permanent pastures) to resist its dispersive effects. In other soils, however,

application of potassium fertilizer can increase soil aggregate stability and potassium nutrition

and yield of irrigated winter wheat, corn, sugar beet and potatoes. Furthermore, in soils of

relatively high organic matter content and higher pH (e.g. 8.2), Koenigs (1961) found that K-

12

rich aggregates were more water-stable than Mg-rich aggregates. Similarly, Weber and van

Rooyan (1971) concluded that K contributes to structural improvement by reducing the Na/K

ratio, and that exchangeable K seems to have a similar effect to Ca on soil stability if the soil is

acidic and exchangeable K exceeds 7%. Further study on the effect of potassium on soil

structure, Chan et al. (1983) found that at low concentration of K+ on exchange sites, there was

a positive effect on the hydraulic conductivity of sandy soils. However, the hydraulic

conductivity deteriorated at high K+ concentration. In dealing with the effect of potassium on

soil structure, Rengasamy and Marchuk (2011) have introduced the useful concept of ‘cation

ratio of soil structure stability’ (CROSS) which incorporates the differential dispersive powers

of Na and K and the difference in the flocculating effects of Ca and Mg. This work shows that

there is better correlation between CROSS and the hydraulic properties of soils than with SAR

which ignores K+ completely.

1.2.2.2 Organic matter

Organic matter is considered an important agent in maintaining the structural stability of a wide

range of soils such as Mollisols, Alfisols, Ultisols and Inceptisols (Baldock and Nelson 2000).

However, the importance of soil organic matter in soil aggregation tends to be less in some soils

such as Oxisols and Andisols due to the important stabilizing role of hydrous oxides or in

Vertisols which contain sufficient clay with substantial shrink/swell potential (Baldock and

Nelson 2000). In soils where organic matter is an important agent binding mineral particles

together, a hierarchical arrangement of soil aggregates exists in which aggregates break down in

a stepwise manner as the magnitude of an applied disruptive force increases (Tisdall and Oades

1982; Oades and Waters 1991; Oades 1993). In general, there are three levels of soil

aggregation proposed by Golchin et at. (1998): (1) the binding together of clay plates into

packets < 20 um, (2) the binding of clay packets into stable micro-aggregates (20 – 250um),

and (3) the binding of stable micro-aggregates into macro-aggregates (> 250 um). The degree of

aggregation and the stability of soil aggregates generally increases with soil organic carbon

(SOC), clay surface area and CEC. In soils low in SOC or clay concentration, aggregation may

be dominated by cations, whereas in soil with high SOC or clay concentration the role of cations

in aggregation may be minimal (Bronick and Lal 2005).

Soil solution composition and the exchange complex

The composition of soil solution and of the exchange complex are in equilibrium so that a

strong predominance of a particular cation in solution is reflected in its contribution to the

exchange complex. Of the four principal exchangeable cations listed above, potassium has

historically been regarded as the minor one so that sodium, calcium and magnesium are seen as

13

the major cations. The equilibrium between soil solution and the exchange complex is

summarized in the Gapon equation (Bresler et al. 1982):

ESR = kG SAR [1.4]

in which kG is the Gapon constant, ESR is the exchangeable sodium ratio given by:

, [1.5]

and describes the solid exchange complex in which Naexch is exchangeable sodium and CEC is

cation exchange capacity (both expressed as charge per unit mass of solid – e.g. cmol(+)/kg).

SAR is the sodium adsorption ratio given by:

, [1.6]

and describes the soil solution in which [Na+], [Ca2+] and [Mg2+] are the concentrations of the

major cations in mmol/L.

The more commonly used description of the exchange complex, ESP, the exchangeable sodium

percentage, is the percentage of the exchange complex accounted for by sodium:

, [1.7]

and is simply related to ESR as:

, [1.8]

The Gapon equation allows simple predictions of how the exchange complex might change

when soil solution composition changes, for example during irrigation and leaching.

1.2.2.3 Clay swelling and dispersion

Swelling is a process by which soil volume increases with water content and decreases (shrinks)

as water content decreases. The mechanism of swelling in sodic soils is well described by

diffuse double layer theory (van Olphen 1977), which accounts for the spatial distribution of a

diffuse layer of exchangeable cations in the space between negatively charged clay particles.

When clay crystals are in close proximity, their diffuse layers overlap, and the total

14

concentration of the ions mid-way between the particles is greater than that in the soil solution

in which the particles are immersed. The difference in concentrations results in a gradient in

osmotic pressure, which thus draws water in between the particles and causes them to move

further apart (i.e. swell). The double layers are extremely thin (ca. < 10-8 m) and can expand or

compress when the electrolyte concentration of the soil solution decreases or increases,

respectively (Quirk and Schofield 1955; Emerson and Chi 1977; Greene et al. 1978b; Arora and

Coleman 1979). The double layer is generally thinner when divalent cations (e.g. Ca2+, Mg2+)

balance the charge and it is thicker when monovalent cations such as Na+ are involved (Quirk

and Schofield 1955). Sodic soils have an elevated proportion of sodium ions on the exchange

complex (exchangeable sodium percentage, ESP > 6) and because sodium is monovalent it

cannot overcome the swelling forces in the double layer, so the clay particles swell and disperse

(Quirk 1986). In pure sodium montmorillonite (i.e. when the entire permanent charge of the

lattice surface is balanced by Na+ ions), large diffuse double layers occur on all clay surfaces

(Warkentin and Schofield 1962; Shainberg et al. 1971; Shainberg and Letey 1984). In dilute

electrolyte solutions, crystalline swelling can produce ten to twenty times the initial dry volume;

in more concentrated electrolytes, swelling is suppressed because the osmotic potential gradient

between the overlapping counter-ions of clay particles and the soil solution decreases

(Warkentin and Schofield 1962). Swelling of Ca-montmorillonite, by contrast, is limited

regardless of electrolyte concentration. This is because, firstly, Ca2+ largely resides in the Stern

layer (or adsorbed layer) and so the diffuse layer is far less populated. Secondly, the crystalline

swelling is not affected by the concentration of calcium with which the clay is in contact (Quirk

1994). Furthermore, when Ca-saturated montmorillonite platelets aggregate in groups of four to

nine platelets, called ‘tactoids’ (Blackmore and Miller 1961), the tactoids effectively reduce the

surface area of the montmorillonite and make them behave like larger particles – thus the

diffuse double layer can only fully manifest itself on the outside surfaces of the ‘tactoids’

(Sumner 1993) – this reduces overall swelling.

When swelling becomes excessive, this leads to dispersion in which clay particles become

separated and move independently of one another. Dispersion usually occurs in sodic soils

where excessive amounts of exchangeable Na are present. When unconfined clays swell in

water, the presence of Na on the exchange sites allows more water to enter between clay

particles and force them apart. The swelling and consequent dispersion both result in an overall

distribution of very small pores due to pore closure and blockage, which degrades the soil

structure, restricts water movement, and discourages root growth (Rengasamy 1983).

15

The minimum total electrolyte concentration required to prevent clay dispersion is called the

‘threshold concentration’ (Quirk and Schofield 1955). Whatever the ESP, the permeability of an

irrigated soil can be maintained by adjusting the electrolyte concentration of the irrigation water

to keep it above the ‘threshold concentration’. The greater the ESP the larger must be the

threshold concentration in the soil solution to maintain permeability. Furthermore, the threshold

concentration depends on the nature of the clay mineral. In Na/Ca systems of montmorillonite

and illite, Oster et al. (1980) found that increasing Na (at relatively low ESP), significantly

raised the threshold concentration required to maintain permeability. This effect was more

pronounced for montmorillonite than for illite; that is, at the same ESP, illite had a higher

threshold concentration than montmorillonite because attractive forces in illite are often smaller

due to differences in edge-to-face surfaces and their irregular step-like surfaces, which cause

mismatching and weaker net attraction between particles (Oster et al. 1980).

More recently Rengasamy and Marchuk (2011) and Marchuk and Rengasamy (2012) have

introduced the “cation ratio of soil structural stability” (CROSS) which provides a more

complete picture of the role of all four of the common exchangeable cations (Ca, Mg, Na, K) in

soil dispersion.

1.2.2.4 Nutritional effects

High levels of exchangeable sodium, and the frequent accompanying high pH of sodic soils,

also restrict the biological transformations and availability of several essential plant nutrients.

For example, the concentrations of calcium and magnesium in soil solution decline as pH

increases due to the formation of relatively insoluble calcium and magnesium carbonates by

reaction with the soluble carbonates that are the causes of the high pH.

Nitrogen deficiency is also a common problem in sodic soils. Excess sodium on the soil

exchange complex imparts structural instability to the soil and causes poor physical properties.

The infiltration rate of the soil is low so that it has restricted internal drainage. For this reason

the surface soil layers remain nearly saturated for prolonged periods following irrigation or rain

resulting in temporary anaerobic conditions. Under alternate aerobic and anaerobic conditions,

loss of nitrogen inevitably occurs through denitrification and volatilization (Patrick and Wyatt

1964). Van Hoorn (1958) also pointed out that under conditions of poor soil structure, twice as

much nitrogen was needed as under conditions of good soil structure.

Phosphorus deficiency in sodic soils is not generally a major problem. Chhabra (1985) reported

that sodic soils contained high concentrations of extractable phosphorus and that there was a

16

positive correlation between soluble P status and the electrical conductivity. Chhabra et al.

(1981) also observed that crops grown in freshly reclaimed sodic soils did not respond to

applied P fertilizer for 4 – 5 years because of their high available P status. However, the

presence of high concentrations of sodium carbonate and soluble P in sodic soils was associated

with calcium deficiencies due to precipitation of calcium phosphates and carbonates from soil

solution.

Potassium nutrition in sodic soils is not clearly understood yet. Some studies indicate that

increasing soil sodicity causes reduced uptake of potassium by crops (Singh et al. 1979), while

other studies (e.g. Martin et al. (1965) and Chhabra (1985) showed the opposite effect. High pH

and presence of calcium carbonate in sodic soil can lead to micronutrient deficiencies; for

example, iron is often limiting in sodic soils. By contrast, boron and molybdenum are rarely

limiting in sodic soils; in fact, they are often present at toxic concentrations. At high pH and

sodicity, boron and molybdenum are present as highly soluble forms that can be taken up by

plants and accumulate in excessive quantities (Pasricha and Randhawa 1971).

1.2.3 Effects of soil structure on soil strength, aeration, and hydraulic conductivity

Soil strength

As mentioned in Section 1.2.1 above, soil aggregates and their stability have an enormous

impact on soil structure and strength (Marshall et al. 1996; Kay 1990; Horn et al. 1995). When

structure becomes unstable or if smaller aggregates are generated, the packing density increases,

inter-particle contacts increase, pore sizes decrease and the overall strength of the soil at a given

water content increases (Lal and Shukla 2004) Horn et al. (1995) also found that soils with

smaller aggregates have greater tensile strength compared to soils with larger soil aggregates,

and that this increases over time with wetting and drying (Figure 1.3)..

This sort of structural degradation is common in saline/sodic soils, which often develop surface

crusts of high packing density and become hardsetting as they dry (Northcote 1979). Crusts

reduce infiltration, increase runoff and have high strength which can reduce crop emergence and

root exploration. Hardsetting soils are unstable when wet and they slump after cultivation to a

density similar to that before cultivation. Because of high density and low macro-porosity they

have low hydraulic conductivity.

17

Figure 1.3 Change in aggregate tensile strength with aggregate diameter as a function of time (after Lal and Shukla 2004)

Soil aeration

Composition, size, and arrangement of pore space located between aggregates are important

factors contributing to water storage for plants. Plant roots need oxygen, which moves much

more slowly through water than through the gas phase (x 104, Marshall et al. 1996). Therefore,

continuous, air-filled pores are required in the soil root zone. Typically, the requirement for

plant development is for at least 10% of the soil volume to comprise gas-filled pores at field

capacity, and for at least 10% of the gas in these pores to be oxygen (Armstrong 1980; Dexter

1988). For this reason, oxygen supply to roots depends on many complex factors including pore

continuity, tortuosity, size and spatial distribution of air-filled pores. A compacted soil or poorly

drained soil contains a smaller amount of air than a well-structured, drained soil. In well-

structured soils the air content is greater because of the presence of macropores. In general, soil

air content and water content are approximately equal at field moisture capacity for well-

structured soils.

In soils that do not shrink or swell, water drainage and extraction by roots results in greater air-

filled porosity because the total pore volume is relatively constant. In soils having a large

shrink-swell capacity the volumetric air content (cm3 pores per cm3) of soil aggregates between

cracks can remain essentially constant or increase by only a small amount as water is removed.

As a shrinking soil dries and cracks, the cracks may provide well-aerated macropore space but

the large peds between the cracks may still be anaerobic and unsuitable for root growth. For this

0.5 1 1.5 2 2.5 3

100

200

300

400

diameter [cm]

tens

ile st

reng

th [k

Pa]

1 year 2 years 4 years

18

reason, only a small proportion of the total soil volume may be exploitable by roots in

swelling/shrinking soils.

Hydraulic conductivity

Stability of aggregates during wetting, and while wet, plays an important role in supplying

water for plants. Aggregates that break down when they are wet have smaller pores, reduced

pore continuity and increased strength upon drying (Cresswell et al. 1992). These all reduce

root penetration, decrease infiltration rate and hydraulic conductivity, and therefore effectively

reduce the amount of water available for plant extraction (Gupta et al. 1989; McGarry 1990;

Lipiec et al. 1991).

1.2.4 Salt-affected soils and their reclamation

Salt-affected soils refer to those exposed to soluble salts during some stage in their history. The

salts involved may include chlorides, sulphates, carbonates and bicarbonates of sodium,

potassium, magnesium and calcium, although the most common salt is sodium chloride. In

general soil salinity in Australia can be classified into two main types: transient salinity and

groundwater salinity. Transient salinity involves a seasonal accumulation of salts in the root-

zone (Rengasamy 2002) and ground-water salinity relates to upward movement of water and

salt by capillarity from a shallow saline water table. In both cases soils become saline when

enough salt remains in the root zone to adversely affect plant growth. Apart from the osmotic

effect of salts in the soil solution, excessive concentration and absorption of individual ions may

prove toxic to the plants and/or may retard the absorption of other essential plant nutrients.

In cases where much of the salt has left the soil profile yet appreciable amounts of exchangeable

sodium remain, the soil is considered to be sodic but not necessarily saline. The criteria used to

classify soils as being either saline or sodic vary across the world. In the United States, salt-

affected soils are classified on the basis of the characteristics of a saturated soil-paste extract

(Richards 1953). In Australia, soils are classified for salinity and sodicity according to the

nature and texture of the whole soil profile and whether the salts are concentrated in the root

zone or at depth (Northcote and Skene 1972). The different criteria for classification are show in

Table 1.3.

Excess soluble salt in soils keeps the clay in a flocculated state so the structure of these soils is

generally good and tillage characteristics and permeability to water can be even better than

those of some non-saline soils. However, when leached with water of low salinity, some saline

19

soils tend to disperse, resulting in low permeability to water and air, particularly in heavy clay

soils.

Table 1.2 Criteria used to classify salt-affected soils in U.S. and Australia.

Soil class

U.S. (saturated soil-paste extract) Australia

(1:5 soil:water extract) ECe

dSm-1 SAR ESP EC1:5*,b

dSm-1 SARb ESPa,b

Non saline – non sodic < 4 < 13 < 15 < 0.7 < 3 < 6

Saline > 4 < 13 < 15 > 0.7 < 3 < 6

Sodic < 4 > 13 > 15 < 0.7 > 3 > 6

Saline-sodic > 4 > 13 > 15 > 0.7 > 3 > 6

Sources: Richards (1953); aNorthcote and Skene (1972); bRengasamy et al. (1984) *The Northcote and Skene definition of salinity is more complex in that it depends on soil texture and profile composition and is based on the mass % of salt in the soil rather than EC

The effects of sodicity usually appear when soluble sodium salts are leached from the soil

profile, thereby leaving some exchangeable sodium bound to clay particles after displacing

cations in solution. There are two main issues associated with soil sodicity: high pH and poor

soil structure. According to U.S. standards, sodic soils are usually low in electrical conductivity

(< 4.0 dS/m) and high in pH (> 8.2). The principal cause of high pH in sodic soils is usually the

hydrolysis of either the sodic exchange complex or of carbonate ions present in CaCO3, MgCO3,

and Na2CO3. The deterioration of structure in sodic soils is caused by a high proportion of

exchangeable sodium on clay-exchange sites, which weaken the bonds between soil particles

when the soil is wetted. This causes clay particles to swell and disperse. Three main problems

associated with soil swelling and dispersion are: reduced infiltration, hydraulic conductivity and

aeration.

The primary cause of the reduction of infiltration is surface crusting. Rain or irrigation water

causes physical dispersion, which results in clay particles becoming mobilised. Soil dispersion

not only reduces the amount of water entering the soil, but also reduces its hydraulic

conductivity because pores close due to swelling and blockage by dispersed particles. Soils in

this state have poor load-bearing capacity (Rengasamy et al. 1984) and they are susceptible to

erosion (Shainberg and Letey 1984; Fitzpatrick et al. 1994). They also have conditions that

generate impenetrable surface crusts that restrict plant emergence and root exploration of the

soil, particularly as the soil dries. Soil with well-defined structure, by contrast, contains a large

number of macropores, cracks, and fissures, which allow for relatively rapid flow of water

20

through the soil. If water cannot pass through the soil, the upper layers can become

waterlogged. This results in anaerobic soils which can reduce or prevent plant growth and

decrease organic matter decomposition rate (Shainberg and Letey 1984; Quirk 1986;

Rengasamy and Sumner 1998).

Both swelling and dispersion are governed by the balance between attractive and repulsive

forces that arise from intermolecular and electrostatic interactions between aqueous and solid

phases in the soil (Rengasamy 1983); each will now be considered in the context of reclaiming

saline/sodic soils.

Two important aspects control the effective reclamation of saline/sodic soils: 1) reducing or

removing exchangeable Na by displacement and leaching, and 2) maintaining the electrolyte

concentration at levels that prevent clay dispersion during leaching (i.e. above the threshold

concentration). Removing Na+ involves its displacement by adding Ca2+. Maintaining the

electrolyte concentration involves the gradual reduction of salinity by leaching during this

displacement of Na+ with Ca2+.

These two requirements are usually achieved in the field by applying gypsum (calcium

sulphate), which is moderately soluble, readily available commercially, and relatively cheap

(Sumner 1993). Maintaining sufficient electrolyte concentrations is crucial because it maintains

water intake on which reclamation depends (Oster 1993). For example, McGeorge and Fuller

(1950) found that Cajon soil in Arizona (high ESP) experienced degradation of soil structure

when irrigated with low electrolyte water – they did not understand the threshold concentration

concept outlined later by Quirk and Schofield (1955). An extensive area of highly sodic pasture

soil (ESP 23) in the Riverina District of New South Wales faced the same problem which was

fixed by dissolving gypsum in irrigation water to decrease solution SAR and to increase the

electrolyte concentration above the threshold value (Davidson and Quirk 1961).

The salt used to keep the electrolyte concentration above the threshold need not be calcium-

based, at least in the first instance. For example, Amemiya et al. (1956) reclaimed a saline-

alkaline soil (ESP 37) in the Coachella Valley of California by mixing sea water with water

from the Colorado River to keep the electrolyte concentration above the threshold value – this

maintained soil permeability while the sea water was gradually diluted and calcium introduced,

and this reduced the period required for reclamation considerably.

Adding gypsum to some Australian red-brown earths (Glenloth and Raywood soils) reduced

ESP and clay dispersion; this also reduced soil strength and increased hydraulic conductivity

21

(Greene et al. 1988). Gypsum can also be added strategically, as done by Kamphorst (1990),

who introduced a solution of gypsum down into soil cracks during the dry season to stabilize the

material on the exterior surfaces of soil peds. This retarded crack-closure during the wet season,

which permitted downward transport of water and salts, improved topsoil aeration and vertical

extension of roots.

Aside from using calcium salts to reclaim sodic soils, one can also add organic polymers into

irrigation water to stabilise the structure and permeability (Kamphorst 1990; El-Morsy et al.

1991). For example, when polyacrylamide (PAM) is added to irrigation water, it can reduce

swelling and dispersion in mildly sodic soils at low EC (Aly and Letey 1990) and can also

increase hydraulic conductivity and leaching of sodic soils (El-Morsy et al. 1991; Zahow and

Amrhein 1992).

In the process of reclaiming saline, alkaline and sodic soils, plants can be used at certain stages

to enhance chemical and physical properties (Qadir et al. 1996). For example, in calcareous

soils, rice can be grown under flooded and ponded conditions in places where plenty of good-

quality water is available. Root and microbial growth in flooded conditions raises the

concentration of CO2 and thus increases the solubility of calcium carbonate (Shainberg and

Oster 1978). Other studies have found that growing Kallar grass over several years can improve

soil organic matter content, hydraulic conductivity, porosity, water retention, structural stability

and plant available water, at least in the surface soil (Akhter et al. 2004).

1.2.5 Models of plant available water

1.2.5.1 Historical models

The above-mentioned factors all influence the amount of water that a plant can extract from the

soil. Until relatively recently, however, most of these factors have not been taken into account

in models to calculate PAW. Early use of the term ‘PAW’ referred to the quantity of water in

the soil ranging from a nominal ‘field capacity’ to a nominal ‘permanent wilting point’

(Veihmeyer and Hendrickson 1927). Field capacity referred to the soil water content at which

excess water after a saturating rain or irrigation drained by gravitational force (in the absence of

evaporation) over a period of several days. The concept was agronomic and was only loosely

linked to a soil matric head (Simmonds et al. 1995). In fact, several different matric heads have

been proposed in different parts of the world (Groenevelt et al. 2001). The ‘permanent wilting

point’, PWP, referred to the soil water content remaining in the soil after plants wilted during

22

the day and did not recover even if placed in an atmosphere of 100% relative humidity

(Veihmeyer and Hendrickson 1949).

The range of soil water content between field capacity and permanent wilting point is rather

imprecise and so cannot accurately define the amount of water that plants can extract from the

soil. This concept of PAW relies on an assumption of equally available water between two

critical potentials, field capacity and permanent wilting point. In reality, it is clear that the

energy plants require to extract a unit of water from soil at field capacity is much lower than at

the permanent wilting point. Furthermore, PAW is also influenced by other factors such as soil

aeration, soil strength, hydraulic conductivity, and salinity. One might observe in some soils, for

example that during drying, plants can extract soil water well beyond PWP. In the same soil

when physical conditions are poor, however, water extraction may stop even when the soil is

quite wet, and certainly before it reaches PWP. Richards and Wadleigh (1952) remarked that the

concept plant water availability should involve two notions: the ability of plant roots to absorb

and use water with which it is in contact, and the readiness or velocity with which the soil water

moves into the root zone to replace that which has been taken up by the plant. Hillel (1971) felt

less confident that soil properties alone could be used to predict soil water availability – he felt

that soil water availability could only be assessed accurately using real plants in real soils under

real meteorological conditions.

To integrate soil physical properties associated with plant growth into the concept of PAW,

Letey (1985) introduced the qualitative concept called the Non-Limiting Water Range, NLWR,

which referred to the range of soil water contents across which limitations to plant growth were

negligible. Water uptake by roots was considered to be directly affected by soil physical

conditions, particularly aeration and mechanical resistance. Accordingly the NLWR became

narrower under conditions of poor aeration and high soil strength (Fig 1.4b) relative to that of

soil of good structural condition (Fig 1.4a).

Da Silva et al. (1994) refined this concept to make it quantitative and called it the Least

Limiting Water Range, LLWR. The LLWR merged the classical water contents at FC and PWP

with those at critical limits of soil aeration and mechanical resistance. In wet soils the upper

limit of LLWR corresponded to the water content at FC (matric head = 1 m), but if the

volumetric air content was less than a cut-off value of 0.10 m3/m3 (selected as being important

from historical literature), then the upper limit of LLWR was adjusted downward until the

volumetric air content reached 0.10. At the dry end, water availability was thought to diminish

due to increasing soil strength such that plant roots could not explore the soil to extract the

water; many plant roots cannot grow into soils that have a penetration resistance > 2 MPa

23

(Cockroft et al. 1969). If soil resistance to penetration was sufficiently low for root proliferation

(< 2 MPa) then the lower limit of LLWR corresponded to the water content at 150 m. However,

if penetration resistance was > 2 MPa before the soil dried to 150 m the lower limit of LLWR

was adjusted upward to a water content where the soil resistance = 2 MPa. The effect of

increasing bulk density on LLWR is shown in Figure 1.5.

Although LLWR incorporated some of the physical limitations affecting water availability (e.g.

aeration, strength) it did not deal with other equally important limitations such as declining

hydraulic conductivity and increasing osmotic stress in unsaturated soils. Furthermore, it used

abrupt cut-off points, whereas real plants experience (and respond to) physical and chemical

limitations in a gradual fashion rather than abruptly.

In earlier work, Feddes et al. (1978) introduced a model that used ‘reduction coefficients’

(varying between 0 and 1) to incorporate various physical and chemical limitations affecting

soil water availability. Their ‘reduction functions’ were fairly simple (linear) and did not

account for the complex nature of plant responses to soil limitations.

Eliminating the complexity of soil limiting factors (e.g. soil strength, aeration, hydraulic

conductivity) that affect plant ability to extract water from soil, Minasny and McBratney (2003)

introduced the concept of integral energy which focuses on the quantity of energy required by

the plant to remove a unit amount of water from the soil. The energy was calculated from the

integral of the soil water retention curve. This work indicated that the energy required to remove

the same amount water from a silty clay soil (within the range FC and PWP) was almost 1.5

times higher than that of clay soil (Minasny and McBratney 2003). The integral energy concept

is important in terms of plant physiology as it considers the plant energy requirements for water

uptake.

24

1.4(a)

1.4(b)

Figure 1.4 The non-limiting water range (NLWR) of water contents as influenced by restricting soil factors for plant growth in soil with (a) good structure and (b) poor structure (Letey 1985).

Figure 1.5 Effect of increasing bulk density on the water content at which volumetric air content = 0.10m3/m3 and soil resistance = 2 MPa, superimposed on the water contents at FC and PWP (after da Silva et al.(1994)); shaded area represents LLWR.

Water content

Fiel

d ca

paci

ty

Perm

anen

t w

iltin

g po

int NLWR

Unavailable water due to rapid drainage (textual limitation)

Water here is held in pores too small to be extracted by most plants (textural limitation)

Water content

Perm

anen

t w

iltin

g po

int NLWR

Water here is unavailable due to mechanical (textural and structural limitation)

Unavailable water due to poor aeration (textural and structural limitation)

Fiel

d ca

paci

ty

A NOTE:


25

1.2.5.2 Integral Water Capacity (IWC) model

Groenevelt et al. (2001; 2004) presented a more flexible model than that of Feddes et al. (1978),

which they called the Integral Water Capacity, IWC. Their ‘reduction coefficients’ consisted of

graded weighting functions intended to mimic plant response to poor aeration, high soil

strength, high/low hydraulic conductivity, and high osmotic stress. They calculated the

differential water capacity, C(hm), as the modulus of the slope of the water retention curve,

�(hm), then multiplied it by appropriate weighting functions, ωi(hm). Each weighting function

accounted for a limiting soil physical property and thus reduced C(hm) to an ‘effective water

capacity’, which was then integrated to obtain the IWC:

� � mmmi

n

idhhChIWC �

��

��

0 1�

[1.9]

where ωi(hm) are multiplicative weighting functions (indicated by the upper case Pi sign, Π)

accounting for up to n limiting properties that vary with the absolute value of the matric head,

hm, in units of cm or m. Examples of how the hydraulic conductivity and poor aeration might

reduce C(hm) to an ‘effective’ water capacity, E(hm), are shown in Figure 1.6 (Groenevelt et al.

2001), which illustrates that the inability of plant roots to take up water at the wet end can be

due to either rapid drainage (because of excessively large saturated hydraulic conductivity in

wet soil) or lack of sufficient aeration (Veihmeyer and Hendrickson 1949; da Silva et al. 1994).

Examples of how osmotic stresses might reduce C(hm) are shown in Figure 1.7 (Groenevelt et

al. 2004), which suggest IWC declined significantly when the EC of a saturated loamy sand

increased.

Figure 1.6 Differential water capacity, C(hm), for the wet end (—); effective differential water capacity, EK(hm), when a hydraulic conductivity weighting function ,ωK(hm), is applied (– –); effective differential water capacity, EKa(hm), when both ωK(hm) and a aeration weighting function, ωa(hm), are applied(Groenevelt et al. 2001).

A NOTE:


26

The examples shown in Figure 1.7 took no account of the physical conditions that deteriorate in

salt-affected soils, so they represent the largest possible amount of water available in a saline

soil. In reality, things would be much worse because the physical properties of salt-affected

soils are generally very poor. Furthermore, these properties vary in a complex way when saline

soils are reclaimed.

Comparing the applicability of the IWC and LLWR models, Asgarzadeh et al. (2010) concluded

that there is a significant correlation between the IWC and LLWR models. However, the IWC

approach relates better to natural phenomena (Asgarzadeh et al. 2010).

Figure 1.7 Weighted differential water capacities for a loamy soil accounting for salt-free conditions and for conditions where EC of the saturated soil = 1, 2, 4, and 7.2 dS/m (Groenevelt et al. 2004).

1.2.5.3 Example of IWC calculations

To evaluate the combined effects on plant available water of soil salinity, limited soil aeration,

and high soil strength (as well as other factors) one can use the procedures suggested by

Groenevelt et al. (2001; 2004). As an illustration, the water retention curve (and fitting

parameters) for Soil 2 (a loamy sand) as given in Fig 1 and Table 2 of Groenevelt et al. (2004)

will be chosen to construct some weighting functions for different soil physical limitations.

The water retention model, �(hm), proposed by Groenevelt et al. (2004) was anchored at two

different points: the saturated point (��= �s, hm = 0 m) and the wilting point (��= �150, hm = 150

m) as described in the relation:

A NOTE:


27

��

�

��

��

��

� ��

�� n

mnm h

kkkh 001150 exp

150exp)( �� , [1.10]

where n and k1 are dimensionless fitting parameters, and k0 is a fitting parameter with

dimension (metre)n. A graphical representation is given for this soil in Figure 1.8.

Differentiating Equation [1.10] gives the basic differential water capacity, C(hm), which defines

the extraction of water per unit change in the matric head:

��

��

� ��

��

��

nm

nm

mm h

khkkndhdhC 01

10 exp)( � [1.11]

It was noted by Groenevelt et al. (2004) that in the presence of salt, Equation [1.11] should

really be known as the differential SOLUTION capacity rather than the differential WATER

capacity because it takes no account of the salt. To account for any salt in the soil solution, a

more general differential water capacity is required, and this is developed below.

Groenevelt et al. (2004) defined the osmo-matric head as:

�omomom hhthhh ,,�� , [1.12]

where �(t, hm,ho) varies between 0 and 1 and is a plant-specific function of t = time, hm and ho,

and is sometimes known as the reflection coefficient (Zimmermann et al. 2002). For the

purposes of this explanation, the reflection coefficient will be set to � = 1, which implies that

plants experience the full effect of any solutes present in the soil solution.

28

��

��

��

omomom dh

dhC �)(

��

��

��

��

��

��

��

��

��

om

mm

om

m

mmom dh

dhhCdhdh

dhdhC )()( �

Figure 1.8 Representation of the water retention curve using data and model for a loamy sand published in Groenevelt et al. (2004).

Equations [1.11] and [1.12] allow one to define the soil water capacity in terms of the total or

‘osmo-matric’ head, hom,:

. [1.13]

One can apply the chain rule to Equation [1.13] to re-define the soil solution capacity, Com(hom),

in terms of the more useful matric head, Com(hm):

. [1.14]

Equation [1.14] suggests (as a first approximation) that a relevant weighting function

accounting for the osmotic effects of solutes is simply the rate of change in the matric head per

unit change in the osmo-matric head, dhm/dhom, viz.

om

mmo dh

dhh �)(� . [1.15]

A NOTE:


29

ECho 6.3

os

o

s

s

hh

ECEC

��

150� om hh

The quantity in Equation [1.15], dhm/dhom, however, is not easily measured, so it must be

separated into its osmotic and matric components. Recalling that for this exercise, � = 1

(Equation [1.12]), inverting Equation [1.15] and separating its parts gives:

� 1111 ��

��

��

��

�

��

� ��

�

��

� ��

�

��

�

m

m

m

o

m

mo

m

mo

m

omo dh

dhdhdh

dhdhdh

dhhhd

dhdh

� , [1.16]

which reduces to

1

1�

��

��

��

m

oo dh

dh� . [1.17]

To calculate the amount of plant available water held in soil using this relation, one assumes as

a first approximation that plants exclude all salt like perfect osmometers (i.e. � = 1) and also

that all solutes are retained in the solution as the water content drops. One can also adopt the

approximate relation between the electrical conductivity of a dilute salt solution, EC, and the

osmotic head, ho, proposed by Richards (1953):

[1.18]

where the unit of EC is dS m-1, for ho is m and the units for the constant are m2 dS-1. With this

information one can calculate a mass balance for the soil solution as follows:

[1.19]

where �s, ECs and hos are the water content, electrical conductivity and osmotic head measured

when the soil is completely saturated (roughly similar to what might be found in a saturated

paste extract). This implies that the product of the water content and the osmotic head at any

point during drying has a constant value which is always equal to the product of the water

content and osmotic head measured at saturation. It also implies, of course, that when plants

wilt, the sum of the two component heads must equal a total of 150 m,

[1.20]

Groenevelt et al. (2004) showed that the following weighting function (to account for the effects

of solutes) can be derived in terms of the matric head, hm, and the osmotic head in the saturated

soil, hos:

30

�

12

001

1

0110 exp

150expexp),(

��

��

��

�

�

��

�

�

��

�

��

��

��

��

��

� ��

��

��

�

��

��

��

� � n

mnwpsosn

m

nmosmom h

kkkhh

khknkhhC ��

1

01

012

001150

exp

exp150

exp

1),(

�

�

��

�

�

��

�

�

��

��

� �

��

�

�

��

�

�

��

�

��

�

��

��

��

��

� ��

��

� nm

nm

nm

n

sososmo h

kh

nkk

hkk

k

hhh

�

�� [1.21]

which can be reduced to:

�1

21),(�

��

��

��

�� m

sososmo hChhh��

� . [1.22]

When Equation [1.22] is substituted into Equation [1.9], the combined water capacity,

Com(hm,hos), can be described in terms of both the matric head, hm, and the osmotic head at

saturation, hos, as follows:

[1.23]

A graphical representation of the combined water capacity for the salt-free condition is shown

by the blue line in Figure 1.9. This corresponds to the differential water capacity derived

straight from the water retention curve of Figure 1.8. The solid red line shows the attenuated

water capacity curve that occurs when salts are present (in this case where the osmotic head in

the saturated soil hos = 2 m). This solid red line shows the most severe attenuation of the water

capacity because it assumes that plants exclude all solutes from the soil solution. The

attenuation would not be so severe in reality where plants allow some solutes to cross root cell

membranes. In this case, however, the wilting point defined in Equation [1.20] is reached when

the matric head, hm = 148 m (i.e. osmotic head, ho = 2 m).

If the section of Figure 1.9 from a nominal field capacity of hm = 1 m into the plant available

range (the region inside the dashed black ellipse) is examined, the effects of salinity plus other

limiting factors (e.g. poor soil aeration, high soil penetration resistance) on the effective water

capacity can be evaluated, and thus determine a theoretical amount of available water using

Groenevelt et al.’s (2004) Integral Water Capacity, IWC, shown in Equation [1.9]. Figure 1.10

shows a segment of the differential water capacity (between hm = 1 and 10 m) from Figure 1.9

for the soil containing no salt (same solid blue line No.1). The solid red line No.2 in Figure 1.10

is the same as in Figure 1.9. The other two (dashed) line segments in Figure 1.10 were derived

as follows.

31

Firstly, plants require oxygen to function normally and it has been postulated that this

corresponds with a minimum volumetric air content in the soil of �air = 0.1 m3 air per m3 soil

(Grable 1966) . If a soil has less air than this, water uptake slows down or stops altogether. One

can thus attenuate the differential water capacity using a weighting function that equals zero for

all matric heads where �air < 0.1 m3 m-3 and equals unity (i.e. no attenuation) for all matric heads

beyond which �air exceeds say 0.20 m3 m-3. The water retention curve shown in Figure 1.8 can

be used to identify the two points where attenuation should begin and end: (�a = 0.10 m3 m-3

when � = 0.357 m3m-3 at hm = 0.5108 m) and (�a = 0.20m3 m-3 when � = 0.257 m3m-3 at hm =

1.4148 m).

Thus the weighting function follows the restrictions:

Figure 1.9 Differential water capacities for the loamy sand of Figure 1.8 when the soil is salt-free (solid blue line) and when the soil has an osmotic head of 2 m in its saturated state (dashed red line). The dotted ellipse identifies the section of the curves discussed in Figure 1.10. ��(hm) = 0 , hm < 0.5108 m, 0 ≤ �(hm) ≤ 1, 0.5108 m ≤ hm ≤ 1.4148 m, [1.24] ��(hm) = 1, hm > 1.4148 m.

32

Figure 1.10 Differential water capacities for salt-free soil (solid blue line, 1), Saline soil with hos = 2 m (solid red line, 2), Saline soil with poor drainage (dashed red line segment, 3), and Saline soil with poor drainage and high strength (dashed purple line segment, 4). For the sake of illustration, a suitable form for this weighting function could be, �air(hm):

� ��

��

�

%10

loghhch m

mair� , 0.5108 m ≤ hm ≤ 1.4148 m, [1.25]

where h10% is the matric head at which the �air = 10% (or �a = 0.1 m3 m-3), and the dimensionless

fitting parameter, c, has a value of 2.262 for this soil. A graphical representation of this

weighting function is shown in Figure 1.11. Substituting Equation [1.25] into Equation [1.9]

along with the effects of salinity produces a small attenuation of the differential water capacity,

shown by the dashed red line segment (3) in Figure 1.10.

33

Figure 1.11 Weighting function to attenuate the water capacity for the effect of poor soil aeration between the matric heads of hm = 0.51 to 1.41 m.

Now if we consider the effects of high soil penetration resistance on soil water availability as

the soil begins to dry, we can explore the effect of restricted root exploration of the soil with a

weighting function, as follows. For the sake of demonstrating things, assume the soil

penetration resistance in the moist soil is low enough to allow full root exploration until the

matric head reaches hm1 = 2 m. As the soil dries beyond hm1 = 2 m, the penetration resistance

rapidly increases until it becomes so great that all root exploration stops completely at a matric

head of say, hm2 = 5 m. A weighting function would therefore equal unity at h m1 = 2 m and

declines to zero at h m2 = 5 m. Thus the weighting function follows the restrictions:��

�(hm) = 1 , hm < 2 m, 1 ≥ �(hm) ≥ 0, 2 m ≤ hm ≤ 5 m, [1.26] ��(hm) = 0, hm > 5 m.

A weighting function for this purpose can take the following form, �SR(hm):

�

�

�

��

�

�

��

�

�

��

�

��

��

��

��

��

�

��

��

��

��

2

2

1

2

2

1

1

m

m

m

m

mSR

hh

hh

h [1.27]

34

where hm1 = 2 m, hm2 = 5 m, and � is a dimensionless fitting parameter having a value of 3 for

this example. A graphical representation of this weighting function is shown in Figure 1.12. By

substituting Equation [1.27] into Equation [1.9] along with the other restrictions (salinity and

poor aeration), the water capacity becomes totally restricted and reduced to zero, as shown by

the dashed purple line segment (4) in Figure 1.10.

Figure 1.12 Weighting function to attenuate the water capacity for the effect of increasingly high soil penetration resistance between the matric heads of hm1 = 2 m, hm2 = 5 m.

Finally, we can integrate all of these restrictions on the water capacity using Groenevelt et al.’s

(2004) Integral Water Capacity (IWC) shown in Equation [1.9]. The results are summarized in

Table 1.2 and they represent hypothetical amounts of water available to plants. The extent to

which they are reasonable estimates has yet to verified. To do this requires extensive plant

response data, which are not yet widely available in the literature (a small data set was

assembled and published in the final stages of preparation of this thesis by Bazihizina et al.

2012).

35

Table 1.3 Summary of physical restrictions on the differential water capacity and their effect on IWC.

Soil restrictions Relevant integrals IWC

m3m-3 Line in Fig1.10

Unrestricted: Non-saline, well aerated, soft structure

� mmom dhhC�1

00,

� mmom dhhC�150

10,

� mmom dhhC�

1500,

0 0.237 0

solid blue

Saline but well aerated, soft structure

� mmom dhhC�1

02,

� mmom dhhC�95

12,

� mmom dhhC�

952,

0 0.154 0

solid red

Saline + poorly aerated but soft structure

� mmom dhhC�1

02,

� � mmommmommair dhhCdhhCh �� 95

4148.1

4148.1

12,2,)(

� mmom dhhC�

952,

0 0.151 0

dashed red

Saline + poorly aerated + hard structure

� mmom dhhC�1

02,

� � � � � mmommSRmmommmommair dhhChdhhCdhhCh �� 5

2

2

4148.1

4148.1

12,2,2,

� mmom dhhC�

52,

0 0.049 0

dashed purple

1.2.6 Conclusions

The literature suggests that plant available water, PAW, is influenced by a plethora of

interacting soil factors including the soil texture, the salinity of the soil solution, the soil

aeration and strength plus the unsaturated hydraulic conductivity. In this review, hypothetical

weighting functions and real soil data have been combined to predict the effects of each of these

factors on soil water availability using the integral water capacity, IWC, outlined by Groenevelt

et al. (2001; 2004). The model of the IWC considers a given soil to be a capacitor that can

release water in a graded fashion according to the severity of various limiting soil physical

properties.

With enough information, it is theoretically possible to calculate the amount of water a given

plant species can extract from the soil under a given set of environmental conditions. To be

universally useful, however, such theoretical calculations must be based on real plant behaviour

– otherwise they would be no better than the simple estimates of PAW proposed more than 50

years ago (Veihmeyer and Hendrickson 1927; Gardner 1960) nor would they offer an advance

36

beyond the linear approach to water extraction offered by Feddes et al. (1978). There are

numerous reports in the literature showing plant responses to various environmental stresses,

but almost none of them is linked quantitatively to soil hydraulic properties.

Furthermore, there is little information describing how plants respond to changes in water

availability, which occur in soils being reclaimed from the saline/sodic state. In the first

instance, there is little commercial interest and expertise to manage the full range of crops that

must be involved to reclaim low-value land, especially if relatively low-value halophytic plants

are needed initially. In saline soils, for example, the concept of plant available water for

Saltbush or Kallar grass might be considered meaningless in a commercial sense because water

use efficiency is not important for these crops. They are simply used to “kick-start” subsoil

drainage and to increase soil organic matter content so that leaching can occur and subsequent

(more valuable) crops can be introduced as part of a plant succession scheme. Furthermore as

salt is leached from the soil profile the physical properties of the soil change dramatically, such

that any new crops that replace the halophytes must deal with a very different set of soil

physical conditions.

Reclamation involves several steps and stages, and although it is broadly understood that the

electrolyte concentration and cation suites must be managed carefully to maintain permeability,

little is known about the changes that occur in soil structure and how these influence plant

available water at each stage of reclamation. A detailed study of soil water availability is

therefore required for a variety of crops at all stages of the reclamation process for a range of

different saline/sodic soils.

1.3 Overall problem, research questions and hypotheses.

1.3.1 Research questions

This study was designed to answer following questions:

1. When soil factors such as strength, hydraulic conductivity, aeration and salinity are taken into

account, how much water is available to nominally ‘salt-sensitive’ plants as predicted using the

Integral Water Capacity (IWC) model of Groenevelt et al. (2004)?

2. For in situ saline field conditions, to what extent does the amount of plant-available water

(predicted using the IWC) match the amount of water extracted from real soils by real plants?

37

3. How does the soil water availability change (as measured using IWC) during various stages

of reclamation of saline/sodic soil?

4. What is the shape of the weighting function describing plant response to increasing salt

concentration in soil? In particular: Is the bent-stick model too simple and is the shape of the

weighting function similar for all plants?

1.3.2 Hypotheses

1. The range in shape, magnitude, and integration limits of the relevant weighting functions (for

strength, aeration, hydraulic conductivity and salinity) will have minimal impact on the amount

of water available. This is because natural variation in the water capacity will be greater than the

variation caused by the relevant soil factors. The alternative hypothesis is that the relevant soil

factors will cause significant variation in the weighted water capacity because their effects apply

across a wide range of soil matric heads from 1 to 150 m and may even overlap. Thus a

significant reduction in the water capacity will be encountered somewhere across the relevant

plant-available range.

2. Extraction of water by plants in the field will match that predicted in the laboratory to be

‘available’ using the IWC. The alternative hypothesis is that the amount of water extracted in

the field will exceed or underestimate that predicted in the laboratory by the IWC, presumably

because certain factors have not been taken into account.

3. Soil water availability (as measured by the IWC) will increase monotonously from one

critical state to the next in soils being reclaimed from the saline/sodic state. The alternative

hypothesis is that the IWC will increase non-monotonously, depending upon the relative

importance of various soil physical properties at each stage of reclamation.

4. A weighting function to attenuate the water capacity for osmotic stress is best described by

the soil-based model of Groenevelt et al. (2004). The alternative hypothesis is that other models,

based upon plant response to salinity are better suited for such attenuation.

38

Chapter 2 Variation in soil water availability down the profile of a saline soil using the

Integral Water Capacity (IWC) model

2.1 Introduction

In Chapter 1, the factors that limit the amount of soil water that is available to plants were

explored and placed in the context of Groenevelt et al.’s (2001; 2004) model, the Integral Water

Capacity, IWC. Three particular soil factors that limit water availability to plants (i.e. salinity,

waterlogging, and high root penetration resistance) were chosen to illustrate the attenuating

effects on the differential soil water capacity of a soil of loamy sand texture. Hypothetical

weighting functions for each limitation were designed (using no real data) simply for illustrative

purposes. The amount of soil water available to plants in the field also depends upon the type of

plant and the variations in soil properties that occur with depth, because plants expand or limit

their root systems to take advantage of better conditions when encountered. Any estimate of soil

water availability for a given crop must therefore consider how the IWC varies down the soil

profile.

There are also other limiting factors to consider in calculating the amount of soil water available

to plants, such as the unsaturated soil hydraulic conductivity. Richards and Wadleigh (1952), for

example, stated that soil water availability involves two dynamic factors: “the ability of the

plant root to absorb and use the water with which it is in contact” as well as the “readiness with

which the soil water moves in to replace that which has been used by the plant”. Soil hydraulic

conductivity differs with texture, structure and tortuosity and of course soil water content or

matric head (Hillel 1982). Furthermore, the hydraulic conductivity of the soil can be too large,

such that excessively rapid drainage removes water from the root zone before it can be used by

plants. For example, Wesseling et al. (2009) found that the amount of plant available water in

coarse-textured soils increased greatly when organic matter was added because it retained water

in the root zone for longer and thus reduced the near-saturated hydraulic conductivity.

The work reported in this chapter describes the procedures used to calculate the variations in

IWC down the soil profile for a compacted, saline, texture-contrast soil. It presents data and

appropriate weighting functions for the limiting soil properties found at a field site (e.g. high

soil strength, low unsaturated hydraulic conductivity, poor soil aeration and high salinity). The

data and functions presented here will form the basis for a comparison with what happens after

land reclamation occurs (covered in subsequent chapters of this thesis).

39

2.2 Materials and Methods

2.2.1 Site selection and sample collection

A survey was conducted at the Roseworthy campus of the University of Adelaide on 15

December 2008 during which soil samples were collected in several paddocks (- 34.52892,

138.67954) to a depth of 20 cm, on which the electrical conductivity of 1:5 soil:water

suspensions was measured. A patch of non-productive, relatively saline, calcareous land was

located in Central-1 (C1) paddock (Figure 2.1) which had an EC1:5 of 0.86 - 0.93 dS m-1 in the

sandy surface horizon. A soil profile was exposed to a depth of 150 cm using a backhoe on 5

January 2009 (Figure 2.2) and the soil classified to the Family level using the Australian Soil

Classification (Isbell 2002) based inter alia upon descriptions of soil colour, texture and

structure (Table 2.1). Nine separate horizons were identified and the soil was classified as an

Epihypersodic Pedal Hypercalcic Calcarosol with Family codes: [C] for the thickness of soil

above Bk horizon, [H] for the gravel content of A1 horizon, [K] for the surface soil texture, [M]

for the maximum texture of B horizon, and [X] for the depth of the whole soil profile.

Disturbed and undisturbed soil samples were collected from the entire soil profile.

Approximately 30 kg of disturbed soil was collected from each horizon and placed in 20 L

buckets for transport to the laboratory for characterization and later use. Three undisturbed soil

cores1 were collected from each horizon by inserting stainless steel rings of dimensions: 50 mm

x 50 mm, vertically into each horizon (Figure 2.3); these were carefully dug out within larger

clods of soil and packed in sealed, insulated containers for transport to the laboratory. The

samples were subsequently used to measure the saturated hydraulic conductivities, water

retention curves, and penetration resistance curves.

Disturbed (bulk) soil samples

The disturbed soil from each horizon was air-dried in the laboratory, passed through a 2 mm

sieve (with gravel content measured), mixed well, and stored again. Subsamples (approximately

0.5 kg) were used to determine particle size distribution, particle density and organic carbon

content, as well as pH, EC and solution cations (Table 2.2). Soil physical properties such as

particle size distribution and particle density were determined by a pipette method (Day 1965)

and pycnometer method (Blake 1965), respectively; organic matter content was determined by

the Walkley and Black method (Allison 1965). pH and EC were measured using 1:5 soil

1 Actually, 5 soil cores were taken in the field but only the ‘best’ 3 were prepared for analysis. The ‘best’ 3 of the 5 were selected by inspection as those having no gaps, cracks or other disturbances.

40

suspensions (Rayment and Higginson 1992); soluble cations were determined on the basis of

saturated paste extract (Janzen 1993). The values presented in Table 2.2 are an average of two

replicates.

Figure 2.1 Roseworthy paddock C1 Figure 2.2 Exposed soil profile. (- 34.52892, 138.67954)

Figure 2.3 Collecting undisturbed soil cores down the profile in paddock C1

41

Table 2.1 Field description of soil physical properties down the profile, plus gravel, particle size analysis and textural triangle description.

Depth, cm Texture by hand Soil colour (dry) Structure Gravel

>2mm%

0 - 10 Loamy sand 10YR 3/1 Very dark gray Loose granular 9

10 - 25 Sandy clay loam 10YR 4/4 Dark yellowish brown Dense platy 16

25 - 35 Light clay 10YR 5/4 Yellowish brown Sub-angular blocky 24

35 - 55 Light clay 7.5YR 6/2 Pinkish grey Sub-angular blocky 24

55 – 75 Light clay 7.5YR 6/6 Reddish yellow Sub-angular blocky 14

75 - 100 Light clay 7.5YR 8/2 Pinkish white Sub-angular blocky 10

100 - 115 Medium clay 7.5R 7/6

Reddish yellow Sub-angular blocky 8

115 - 150 Medium clay 5YR 6/3

Light reddish brown Angular blocky 4

> 150 Heavy clay 2.5YR 4/6 Red Sharp angular blocky 0

Table 2.2 Bulk density of undisturbed soil cores, �b, and particle density, �s, particle size analysis (international system), pH, EC, organic matter content, OM, and SAR from bulked soil samples in each horizon.

Depth, cm

��b g cm-3

�s g cm-3 Sa

nd %

Silt

%

Cla

y% Texture

from particle size

pH1:5 ECe*

dS m-1 OM % SAR

0-10 1.05 2.37 65 19 15 Loam 7.42 6.29 10.29 2 10- 25 1.58 2.58 50 24 27 Clay loam 8.24 7.63 1.52 5 25- 35 1.52 2.64 49 23 28 Clay loam 8.30 6.35 0.76 7 35- 55 1.54 2.68 47 24 30 Clay loam 8.47 6.03 0.54 8 55-75 1.54 2.69 43 25 32 Clay loam 8.91 5.33 0.05 14 75-100 1.57 2.73 41 26 33 Clay loam 9.51 4.36 0.09 30 100-115 1.63 2.72 35 30 35 Silty clay

loam 9.48 4.71 0.00 37

115-150 1.69 2.74 35 26 39 Clay 9.34 5.68 0.00 39

> 150 1.73 2.72 33 20 47 Clay 8.86 6.66 0.00 42 * EC and pH were measured in 1:5 soil:water extracts; EC1:5 values were then transformed to ECe values using the method of Slavich and Petterson (1993). EC was measured first, then 0.01 M CaCl2 was added, samples re-shaken, and pH measured.

42

Undisturbed soil cores

The undisturbed soil cores from each horizon were carefully removed from their sealed,

insulated containers, trimmed at both ends, and a piece of 7 x 7 cm porous material (38 �m

mesh) fitted tightly over one end. The soil cores were then wetted by capillary action using

isotonic solutions prepared using the cation concentrations and SAR values determined on the

bulked soil samples (Table 2.2). Measurements of saturated hydraulic conductivity, water

retention, and penetration resistance were taken on each soil core using isotonic solutions as

follows.

2.2.2 Saturated hydraulic conductivity, water retention, and soil penetration resistance.

The hydraulic conductivity of each saturated soil core was measured by fastening an extension

to the top of each ring (clamped to a retort stand) to allow a hydrostatic head of 4 cm to be

(gradually) established and maintained by a constant head device (an inverted 1-litre bottle

containing the isotonic solution applicable to each soil horizon (Figure 2.4). A circular piece of

Whatman No.2 filter paper was placed on the surface of each soil core before the hydraulic head

was established to minimise surface disturbance. The flux of solution was monitored hourly

over several days until it reached steady state, at which time the hydraulic conductivity was

determined from the steady state flux and the hydraulic gradient.

Figure 2.4 Laboratory set-up to measure saturated hydraulic conductivity on undisturbed soil cores prior to measuring their water retention curves using field-isotonic solutions.

43

Each sample was then unclamped from its retort stand, its extension removed, and placed onto a

porous ceramic plate (saturated with an identical isotonic solution) held at a progressive series

of matric potentials to determine the water retention curve. Hanging columns of isotonic

solutions were used for the smaller pressure heads (e.g. 2.5 cm to 100 cm)2 soil cores were

exposed to these pressure heads successively over a period of 2 days (for h = -2.5 cm) or up to 7

days (for h = -100 cm). For greater pressure heads, sealed, high-pressure chambers (supplied

with pressurized N2 gas) were used for the following periods: 7 days for h = -500 cm, 15 days

for h = -1,000 cm, 25 days for h = -5,000 cm, 45 days for both h = -10,000 cm and -15,000 cm.

At each pressure head, samples were removed from their ceramic plates, weighed and placed

back onto their pressurized plates at the same pressure heads, with additional water added to the

soil samples to ensure good soil-plate connection, for a further 48 hours. This was repeated until

the weights did not change by more than 0.1% between readings, which ensured the samples

were as close to equilibrium as possible. The equilibrium weights at each pressure head were

recorded and the volumetric water contents calculated at the end of experiment after the oven

dried soil masses (105oC for 48 h) were measured and the bulk volumes checked.

After the final weight of each soil sample was recorded for each pressure head, the penetration

resistance was measured using a Lloyd Instrument LF-plus Penetrometer connected to a

computer with data-logging software NEXYGENPlus to record the penetration force

encountered by an 85 mm long stainless steel pin with 30o cone angle; 2.58 mm diameter cone-

tip (diameter of the recessed shaft behind the cone was 1.95 mm). The cone was inserted

vertically at a constant rate (2.8 mm/min) and the force recorded every 0.5 mm to a depth of 45

mm below the soil surface. The penetration resistance (force encountered divided by the cross-

sectional area of the cone) was calculated as the average value between the depths of 15 to 35

mm; this minimized surface effects at the top and bottom of cores.

2.2.3 Salinity and osmotic stress

As indicated in Chapter 1, the weighting-analysis for osmotic stress proposed by Groenevelt et

al.(2004) requires information on the salt concentration of the soil in its saturated state, ideally

from a saturated paste extract. In this work, I initially measured EC on 1:5 soil:water extracts,

not paste extracts, so the 1:5 EC values were transformed to paste extract EC using the method

proposed by Slavich and Petterson (1993).

2 Samples exposed to the hanging columns of solution were enclosed in a plastic membrane to reduce evaporation.

44

2.3 Result and discussion

2.3.1 Saturated hydraulic conductivity

As might be expected with increases in clay content with depth in the soil profile, the mean

saturated hydraulic conductivities, Ks (m s-1), declined by over 5 orders of magnitude from the

sandy-textured soil surface down to the heavy clay-textured subsoil at 1.5 m depth (Figure 2.5).

The variability in Ks values among the 3 samples, shown in Figure 2.5 using red horizontal

standard error bars, is surprisingly small for such readings, which are often log-normally

distributed (Hillel 1971). The values of Ks will be referred to later in reference to calculating the

unsaturated hydraulic conductivity function of Grant et al.(2010).

Figure 2.5 Saturated hydraulic conductivities of undisturbed soil cores down the soil profile using isotonic solutions applicable to each depth (horizontal red bars are standard errors.

2.3.2 Water retention curves

The water retention data for all 9 horizons (Figures 2.6) were fitted to Equation [1.10] using a

Levenberg-Marquarde least-squares optimization procedure, the parameter values for which are

given in Table 2.3. The water retention curves were grouped in Figures 2.6 based on their

texture and salt concentration: Figure 2.6a shows curves for the two light-textured horizons in

the top 25 cm, which had EC values ranging between 6 and 8 dS m-1. Figure 2.6b shows the

curves for the next 3 soil horizons (25 to 75 cm), which had medium textures and EC values

ranging between 5 and 6 dS m-1. Figure 2.6c shows the curves for the next two horizons (75 to

Depth, cm Mean Ks m s-1 (± std error)

0 - 10 6 x10-5 (3 x10-5)

10 - 25 3 x10-5 (8 x10-6)

25 - 35 4 x10-5 (6 x10-6)

35 - 55 3 x10-5 (4 x10-6)

55 – 75 2 x10-5 (7 x10-6)

75 - 100 2 x10-5 (3 x10-6)

100 - 115 7 x10-7 (2 x10-7)

115 - 150 2 x10-7 (6 x10-8)

> 150 8 x10-9 (1 x10-9)

45

115 cm), which had greater clay contents and less salt (ca. 4 dS m-1). The curves for the deepest

(heavy clay) horizons (115 to 150 cm) are grouped in Figure 2.6d...

Figure 2.6 Water retention curves for the 9 soil horizons examined in this study: a) 0 to 25 cm, b) 25 to 75 cm, c) 75 to 115 cm, and d) 115 to 150 cm.

a)

d)

c)

b)

46

...and had EC values ranging between 5 and 7 dS m-1.

At a glance, the water retention curves in Figures 2.6 reflect the texture and structure of the

different soil horizons. For example, the sandy surface horizon with very high organic matter

content (Figure 2.6a) had relatively steep slopes across the wet range, which simply reflected

the preponderance of large pores stabilised by organic material (Peerlkamp 1950; Emerson and

Smith 1970; Verma and Sharma 2008). As the texture and bulk density gradually increased and

the organic matter content decreased with depth (creating a narrower distribution of smaller and

smaller pores), the water retention curves became flatter.

Table 2.3 Fitting parameters, k0, k1 and n, for the water retention curves in each horizon down the soil profile. Optimization of the fitting parameters was conducted with fixed (measured) values of the volumetric water content at saturation, �s, and permanent wilting point, �150.

Horizon depths (cm) ��s �150

ko (metre)n k1 n

0 - 10 0.552 0.186 234 0.400 0.564

10 - 25 0.386 0.247 1,409 0.205 0.423

25 - 35 0.421 0.225 153 0.204 0.676

35 - 55 0.425 0.208 341 0.243 0.584

55 - 75 0.426 0.172 333 0.277 0.634

75 – 100 0.421 0.147 2,986 0.520 0.307

100 – 115 0.395 0.242 897 0.222 0.353

115 – 150 0.381 0.277 276 0.126 0.422

> 150 0.363 0.275 747 0.117 0.429

2.3.3 Soil penetration resistance

The soil penetration resistance (MPa) as a function of the soil matric head (cm) is plotted for

each soil in Figures 2.7. The data were fitted to power functions (solid lines) of the form;

SR(hm) = a hmb , [2.1]

where a, and b are adjustable fitting parameters. A Levenberg-Marquardt least-squares

optimization procedure was used in the computer software package, Mathcad 14.0 (Table 2.4).

Of primary interest was whether the soil resistance fell between the horizontal dashed-green line

at 0.5 MPa (when soil resistance begins to restrict root exploration of the soil) and the horizontal

47

dashed-red line at 2.5 MPa (when soil resistance completely restricts all root growth of most

plants in the soil). The critical values of 0.5 and 2.5 MPa were chosen from historical data of

(Greacen et al. 1968) and (Cockroft et al. 1969). The soil matric heads at which penetration

resistances of 0.5 and 2.5 MPa were encountered for each soil were calculated by re-arranging

Equation [2.1], substituting the initial and final values of SR, and solving for hi and hf:

and [2.2] Values for hi and hf are shown in Table 2.4.

Table 2.4 Fitting parameters for Equation [2.1] describing the relation between soil penetration resistance (MPa) and soil matric head (cm), plus the matric heads, hi and hf, respectively, at which SR(hm) reached values of 0.5 and 2.5 MPa.

Horizon depth (cm)

a (MPa cm-b) b hi

(cm) hf

(cm) 0 – 10 0.12 0.26 215 103,900

10 - 25 0.29 0.24 10 8,824 25 – 35 0.05 0.42 221 10,127 35 – 55 0.24 0.25 22 14,094 55 – 75 0.12 0.40 36 2,035

75 – 100 0.06 0.42 146 6,784 100 – 115 0.06 0.46 117 3,828 115 – 150 0.09 0.45 53 1,905

> 150 0.024 0.56 239 4,283

It can be seen in Figure 2.7a that the soil resistance stayed well within the tolerable range across

all pressure heads for the top soil horizon (high organic matter). For the next soil horizon, the

soil resistance only just exceeded 2.5 MPa when the soil matric head dried to hm > 5,000 cm.

For all horizons from 55 cm downward (with greater densities, clay contents and lower organic

matter contents), the soil resistance significantly exceeded 2.5 MPa for hm < 5,000 cm (Figure

2.7 b, c and d). In fact, the soil resistance entered the restrictive range for some soil horizons

when the matric head was only hm = 100 cm.

2.3.4 Salinity and osmotic stress

The measured electrical conductivities were converted to ECe (Table 2.2), which will be used as

the first weighting function to attenuate the water capacity, following Groenevelt et al. (2004).

The measured and corrected EC values, plus the corresponding (calculated) values of the

osmotic head in the saturated soil, hos (cm), and the matric head at which (hos + hm) = 15000 cm

(i.e. the permanent wilting point) are shown in Table 2.5.

48

Figure 2.7 Soil penetration resistance (SR, MPa) as a function of matric head (hm, cm) for the same soils presented in Figures 2.6. The data falling between the horizontal green and red dashed lines represent conditions that increasingly restrict root growth in the soil.

a)

b)

c)

d)

49

Table 2.5. Measured values of the electrical conductivity of 1:5 soil:water extracts and gravimetric water contents at saturation, plus the corresponding electrical conductivity of paste extracts (calculated from Slavich and Petterson (1993)) and values of hos and hm at wilting point (calculated from Equation 12 in Groenevelt et al. (2004)).

Depth (cm)

Measured Calculated EC1:5

(dS m-1) �S

(g g-1) ECe

(dS m-1) hos (cm)

hm (cm)

0 - 10 0.96 0.74 6.29 2264 9420 10 - 25 0.80 0.43 7.63 2747 10708 25 - 35 0.71 0.46 6.35 2286 10883 35 - 55 0.65 0.45 6.03 2171 10879 55 -75 0.57 0.44 5.33 1919 10775 75 - 100 0.70 0.79 4.36 1570 10504 100 - 115 0.85 0.98 4.71 1696 12232 115 - 150 1.09 1.09 5.68 2045 12191 > 150 1.52 1.57 6.66 2398 11835

2.3.5 Weighting functions

2.3.5.1 Weighting the differential water capacities for salinity

The water retention curves, �(hm), shown in Figures 2.6 were differentiated with respect to hm to

produce differential water capacities, C(hm) according to Equation [1.11] after Groenevelt et al.

(2004). These were then transformed to differential soil solution capacities, Com(hm) by taking

into account the salt concentrations from saturated paste extracts3 according to Equations [1.12],

[1.13], and [1.14].

The soil solution capacities were then attenuated according to Equations [1.15] to [1.22], which

produced effective soil solution capacities for each saline condition using Equation [1.23]. The

attenuated curves (dotted lines) are shown together with the non-attenuated curves (solid lines)

in Figure 2.8.

The weighted differential water capacities in Figures 2.8 come from Equation [1.23] and

represent the starting points for all subsequent attenuations to account for other limiting soil

physical properties (e.g. high penetration resistance, poor soil aeration, low unsaturated

hydraulic conductivity). All soil horizons contained soluble salts so all of them experienced

some attenuation of the water capacity – some horizons more severely than others. The effects

3 Saturated paste extract-EC were estimated by correction of the EC1:5 according to the method proposed by Slavich and Petterson (1993).

50

of the different attenuations on the integral water capacity, IWC, will be summarised at the end

of this chapter.

2.3.5.2 Weighting the differential water capacity for high soil penetration resistance.

The weighting function developed here acknowledges that soil strength alone rarely stops plant-

extraction of water from the soil completely – it merely limits the volume of soil into which

roots can grow; soil water can still flow toward roots that enter cracks or biopores present in the

high-strength soil matrix so roots can still take up water. Acknowledging that a weighting

function for high soil resistance, SR(hm), needn’t necessarily end at zero simply because root

extension stops, the magnitude of SR(hm) is more flexible than for other physical restrictions.

The values of SR(hm) were therefore restricted as follows:

�SR(hm) = 1.0 for h < hi, and 0 < �SR(hm) < 1 for hi < hm < hf ,

where hi is the soil matric head at which SR(hm) = 0.5, and hf is the soil matric head at which

SR(hm) = 2.5. The values of hi and hf were calculated using Equations [2.2] and are shown in

Table 2.4.

The form of the weighting function chosen in this study for soil resistance took a more flexible

form than that described in Equation [1.22], and can be adjusted in future work for plants

having different ability to penetrate hard soils, as follows:

[2.3]

where SR(hi) = 0.5 MPa, SR(hf) = 2.5 MPa as described above.

The parameter � in Equation [2.3] is a dimensionless slope-parameter designed to create

different severities of attenuation depending on the ability of different plant species to exert root

growth pressures on their surroundings. For example smaller values of � would be used for

plant species that are able to exert higher root growth pressures, so the attenuation is less severe.

By contrast, a larger value of � would be used for plant species known to be sensitive to

compaction or that cannot exert high root growth pressures; so the attenuation would be much

more severe.

51

Figure 2.8 Differential water capacities for the nine water retention curves shown in Figure 2.1 weighted (dotted lines) or not weighted (solid lines) for salt content according to Groenevelt et al. (2004).

Figure 2.9 shows 3 possible weighting functions for each of the nine horizons by adjusting the

�-parameter from � = 0.2 to � = 0.5 to � = 1.0, respectively, for plants that can exert ‘high’

‘medium’ or ‘low’ root growth pressures on their surroundings. There is no published

information on the magnitude of such �-values, so in this study a value of � = 0.5 was chosen

for illustration purposes. Figure 2.9 shows that the attenuations occurred across a wide range of

different matric heads for the different soil horizons. Some attenuations are more moderate than

others, for example, the surface soil horizon (which contained 18% organic matter) was

attenuated the least, whereas the deeper soil horizons (which had very great strength) were

attenuated more severely. Notably, none of the weighting functions causes complete attenuation

to zero across the range of soil matric heads examined here, which (as explained above) is

probably realistic. As for the effects of salinity, the effects of high soil strength on IWC will be

summarized at the end of this Chapter, but will clearly be more severe in some horizons than in

others.

A NOTE:


52

Figure 2.9 Three possible shapes for weighting functions to attenuate the water capacity based upon the ability of different plants to exert higher or lower root growth pressures on their surroundings. Upper dotted lines come from using � = 0.2 (for strong plant roots), solid lines come from using � = 0.5 (for medium-strength plant roots), and lower dash-dotted lines come from using � = 1 (for weak plant roots).

2.3.5.3 Weighting the differential water capacity for poor soil aeration.

The weighting function for poor soil aeration is based upon the historic literature describing the

minimum volumetric air content, �air, required for normal growth of plants in wet soils. (Grable

and Siemer 1968) and da Silva et al. (1994) suggested �air should be at least 0.1 cm3 air cm-3

total volume, and that plants respond to increased amounts of air up to a volumetric air content

of 0.2 cm3 cm-3. Using this convention, an effective weighting function can be prepared,

ωair(hm), to allow the water capacity to be fully attenuated, ωair(hm) = 0, when the soil is

saturated, and to gradually increase to ωair(hm) = 1.0 as the soil drains and dries out, according to

the relation proposed by Grant et al. (2003):

53

[2.4]

where

hi is the matric head at which �air is the minimum critical value, 0.1 cm3 cm-3.

hf is the matric head at which �air is 0.2 cm3 cm-3.

A is a dimensionless slope parameter designed to create different severities of attenuation

depending on the ability of different plant species to cope with poorly aerated soil conditions.

Large values of A, for example, would be used for plant species that are very sensitive to poor

soil aeration (e.g. tomatoes4); such values would prolong the attenuation until the soil drained

and dried out to a greater extent. By contrast, smaller values of A would be used for plant

species that tolerate poor soil aeration very well and recover rapidly when conditions improve;

such values would see the attenuation removed quickly. Figure 2.10 presents three (among an

infinite number of) possible weighting functions for each of the nine soils by adjusting the A-

parameter from A = 0.2 for plants that easily tolerate poor soil aeration (upper lines), to A = 0.5

for plants that have an average tolerance for poor soil aeration (middle lines), to A = 1.0 for

plants that are very sensitive to poor soil aeration (lower lines). It can be seen for some soil

horizons that aeration quickly disappears as a physical limitation such that the weighting

functions go from 0 up to 1 across a very narrow range of very small suctions (e.g. 0 – 10 cm).

At the other extreme, some soil horizons are completely limited by poor aeration such that the

weighting function retains a value of 0 all across the range of matric heads considered here (e.g.

> 150cm). The other soil horizons have conditions that require weighting functions between

these two extremes.

With no published data on such A-values for different plants, the value A = 0.5 will be used for

illustration purposes in this work. As indicated above for salinity and soil resistance, the effects

of weighting for soil aeration on IWC will be summarized at the end of this Chapter.

4 Literature indicated that optimum growth of tomatoes occurred in soil with air space at about 30 to 35% (Flocker et al. 1959)

54

Figure 2.10 Three possible shapes (of many) for weighting functions to attenuate the water capacity for poor soil aeration by varying the A-parameter in Equation [2.5] from 0.2 (upper dotted lines), 0.5 (middle solid lines), and 1.0 (lower dash-dot lines) according to the ability of different plants to tolerate poor soil aeration.

2.3.5.4 Weighting the differential water capacity for declining soil hydraulic conductivity.

Weighting the water capacity for declining hydraulic conductivity requires knowledge of the

unsaturated hydraulic conductivity function for a soil. A theoretical framework linking the water

retention curve to the hydraulic conductivity function was presented by Grant et al (2010), and it

is possible to use the parameters from this theory to develop an appropriate weighting function

as follows.

55

By scaling the water retention curve, �(hm), using the relative water content, �r = �/�sat, rather

than the absolute water content, �, the water retention model takes the form (Grant et al. 2010):

, [2.5]

where �s is the water content at saturation and k0, k1 and n are fitting parameters that depend on

the shape of the water retention data.

To obtain a relative hydraulic conductivity from Equation [2.5], it needs inter alia to be

integrated. The integrand of interest in this work has the form:

! � " , [2.6]

where ! = �s / k1, and " = 2/n following the Burdine restriction (Grant et al. 2010).

Inserting Equation [2.5] into Equation [2.6] and carrying out the integration yields an expression

that represents the unsaturated hydraulic conductivity as an incomplete gamma function, M(h):

�# $# $# $��

�

�

��

�

�

��

�

�

��

�

�

��

�

�

��

�

�

��

�

�

��

�

�

��

�

�

��

�

�

��

�

��

�

��

�

��

��

��

��%��%

n

msrLm h

kkhM 01 exp11ln,11ln,11)(�

!"�!"!

[2.7]

where �rL in Equation [2.7] is defined as the lower limit of integration (Grant et al. 2010), and

all other variables have been defined above.

To obtain an expression for the relative hydraulic conductivity, Kr(hm), Equation [2.7] is

substituted into the following expression:

, [2.8]

where M(1) is the expression for the hydraulic conductivity at saturation.

With this expression, it is now possible to produce a weighting function that addresses the

problem of declining unsaturated hydraulic conductivity and allows attenuation of the water

capacity from 1 down to 0 at specified matric heads:

56

, [2.9]

where k0 comes from the water retention curve (Equation [2.5]), hi and hf are the matric heads at

which plants initially experience water stress due to low hydraulic conductivity, and when they

finally wilt permanently and die. The matric heads at the onset and finality of water stress due to

declining hydraulic conductivity probably depend on plant species and environmental

conditions, and an investigation of this is beyond the scope of the present study. For lack of any

other information, one can select hi = 2500 cm, which was proposed by Gardner and Nieman

(1964) to be a reasonable average matric head at which water stress due to limited unsaturated

hydraulic conductivity begins to restrict the growth of many plants. One can also select hf =

15,000 cm, which is typically reported as permanent wilting point anyhow.

The parameter, �, in Equation [2.9] is proposed to be a plant-specific slope parameter, for which

small �-values apply to plants sensitive to low hydraulic conductivity and for which large �-

values apply to plants that cope well with declining hydraulic conductivity. In the complete

absence of any published literature on appropriate values for such a parameter, one can select �

= 0.2 as being ‘small’ for sensitive plants, � = 0.5 for plants of ‘medium’ sensitivity, and � = 1

as being a ‘large’ value for tolerant plants.

A set of three weighting functions for declining hydraulic conductivity are therefore plotted in

Figure 2.11 for plants differing in their ability to tolerate such conditions. Obviously they all

start and end at the same place and they all have similar shapes because there is no information

to vary the values of hi, hf and � in Equation [2.9] – this will require experimental work with real

plants under a range of controlled conditions – beyond the scope of this thesis!

2.3.6 Summarizing the effects of weighting the water capacity

For the purposes of demonstrating the effect of weighting the water capacity using the functions

displayed for salt (Equation [1.23]), soil resistance (Equation [2.3]), soil aeration (Equation

[2.4]) and hydraulic conductivity (Equation [2.9]), one can start with the attenuated ‘effective’

water capacity accounting for salt shown in Figure 2.8 and apply the weighting functions shown

in Figures 2.9, 2.10 and 2.11. Because there is no unbiased information to help select the slope-

parameters in the weighting functions mentioned above (i.e. � in Equation [2.3], A in Equation

57

[2.4], and � in Equation [2.9]), only the central, solid lines (middle values for the slope

parameters � = 0.5, A = 0.5 and � = 0.5) will be used in this analysis.

Table 2.6 summarizes the IWC-values obtained as each weighting function is applied to

attenuate the differential water capacity separately for each soil horizon, and also when all

weighting functions are applied simultaneously. The classical, unweighted estimate of plant

available water, PAW, is also included for comparison.

It is clear in columns 2 and 3 of Table 2.6 that the greatest attenuation of the water capacity

comes from accounting for the high salt content of the soil, particularly in the top 100cm. For

example, by weighting the water capacity in this way, the amount of “plant available water”

declines from 284 mm/m in the top 10 cm to only 124 mm/m, a reduction of more than 50%.

Reductions in PAW for the other horizons in the top 100 cm range between 36 and 57%. Below

100 cm in the soil profile, the clay content is very large so water retention in small pores

reduces the amount of water that plants can extract in the first place, so taking salinity into

account only reduces PAW modestly. For example, in the zone 100 to 110 cm, PAW decreased

from 128 mm/m to 89 mm/m (30% reduction) and below this depth, the reductions in PAW

were in the range 27 to 29% due to salt.

Figure 2.11 Three possible shapes for weighting functions to attenuation the water capacity based upon the ability of different plants to cope with declining hydraulic conductivity in dry soils. The lowest dotted lines come from using � = 0.2 in Equation [2.9] for sensitive plants; the highest dash-dot lines come from using � = 1.0 for tolerant plants, and the central solid lines come from using � = 0.5 for medium plants.

58

Tab

le 2

.6 P

redi

ctio

ns o

f pla

nt a

vaila

ble

wat

er in

a sa

line

soil

prof

ile b

ased

upo

n va

rious

deg

rees

of w

eigh

ting

of th

e di

ffer

entia

l wat

er c

apac

ity; i

nteg

rals

at t

he

top

of e

ach

colu

mn

indi

cate

the

type

of w

eigh

ting

appl

ied:

PA

W =

cla

ssic

al a

ppro

ach

with

no

atte

nuat

ion,

IWC

= in

tegr

al w

ater

cap

acity

with

atte

nuat

ions

to

acco

unt f

or, r

espe

ctiv

ely:

salt

alon

e, sa

lt +

poor

aer

atio

n, sa

lt +

high

soil

resi

stan

ce, s

alt +

dec

linin

g hy

drau

lic c

ondu

ctiv

ity, a

nd a

ll fa

ctor

s com

bine

d.

Dep

th

cm

�

�

�

�

PAW

(no

atte

nuat

ion)

IW

C S

alt a

lone

IW

C S

alt +

Aer

atio

n IW

C S

alt +

SR

IW

C S

alt +

K(h

m)

IWC

All

fact

ors

mm

/ m

0-

5 28

4 12

4 10

9 98

12

0 81

5-

10

284

124

109

98

120

81

10-2

0 13

2 84

8

48

74

3 20

-30

137

84

22

57

78

12

30-4

0 16

4 93

47

63

89

26

40

-50

187

101

59

61

95

30

50-6

0 20

4 10

8 72

61

10

5 33

60

-70

221

114

84

61

115

36

70-8

0 23

7 11

1 84

61

10

3 35

80

-90

252

108

83

61

90

34

90-1

00

252

108

83

61

90

34

100-

110

128

89

18

51

78

5 11

0-12

0 10

3 73

11

41

66

3

120-

130

78

57

5 30

53

0

130-

140

78

57

5 30

53

0

59

After salinity is taken into account, the data in Table 2.5 suggest that the approximate order of

importance of the other individual limiting factors was: high soil strength > low soil aeration > low

hydraulic conductivity. There are no statistical boundaries on these estimates so they are only

approximate, but sustained poor soil aeration and high soil strength are known to severely cut the

amount of water plants can extract from the soil, whereas the effect of low hydraulic conductivity

can depend on plant demand (Chahal 2010).

When all factors are taken into account simultaneously, there is a substantial reduction in PAW, in

some cases to absolutely zero. Of particular note is the great reduction in PAW in the layer from

10-20 cm, where the bulk density, degree of saturation and soil hardness were greater than above

and below. If plant roots were unable to exploit cracks and biopores to get through this dense, hard

layer, the plants would surely perish from water stress as the surface soil dried out. This is best

illustrated by the sharp reduction in available water in the 10-20 cm depth (Figure 2.12)

Figure 2.12 Amount of plant available water down the profile of a saline soil (mm/m) predicted by taking into account different soil physical restrictions listed in Table 2.5.

60

2.4 Conclusions

This study attempted to quantify how much water is available to nominally ‘salt sensitive’ plants in

a saline soil using the model proposed by Groenevelt et al. (2004). It was hypothesised that the

relevant weighting functions would have minimal impact on the amount of water available. The

alternative hypothesis proposed was that when the relevant soil factors were taken into account to

attenuate the differential water capacity, a significant reduction in available water would occur

because the effects of the physical limitations might be multiplicative across the normal range of

plant available water.

On the basis of the results present here, the null hypothesis can be safely rejected in this study and

the alternative tentatively accepted: ‘accepted’ because there were ‘clear’ effects of the weighting

functions, but ‘tentatively accepted’ because a statistical evaluation of the effects was not possible

and because many of the weighting functions were applied with little or no knowledge of the real

magnitude of their parameters based upon real plant behaviour. Given the limitations on the

findings in this Chapter, it is essential to obtain real plant responses on the same soil in the field,

which is the topic of Chapter 3.

61

Chapter 3 In situ response of plants to saline conditions in the field

3.1 Introduction

Plants growing under saline field conditions have to deal with both osmotic and specific-ion

stresses (Shainberg and Oster 1978). For example iso-osmotic solutions of various different salts

are known to reduce growth in similar fashions (Bernstein and Hayward 1958). Furthermore, soil

water matric and osmotic stresses have a similar and additive effect on plant growth (Wadleigh and

Ayers 1945) and these both reduce water uptake and transpiration of plants (Meiri and Poljakoff-

Mayber 1970; Hoffman et al. 1971). For the purposes of this work, I will exclude ion toxicities as a

variable in water uptake by using a plant that is relatively tolerant of chloride and sodium.

Salinity and sodicity vary with both the soil water content and with depth, so plant available water

varies down the soil profile and within the root zone of most soils. In the absence of significant

drainage and leaching of soluble salts, the salt concentrations generally increase during drying.

Even when leaching of salts is possible, the salt content at the bottom of the root zone can be

significantly greater than it is at the top, particularly if the leaching fraction drops below about 0.3

(Richards 1953).

As salt concentration declines during leaching, it is well known (Quirk and Schofield 1955) that

soil structure and structural stability begin to change and that these changes influence plant

available water through their effects on soil porosity, aeration, penetration resistance, and hydraulic

conductivity. Degraded soil structure in sodic soils has been reported by many researchers (e.g.

Richards (1953), Rengasamy and Olsson (1991), Jayawardane and Chan (1994)) as reduced

hydraulic conductivity and aeration and as increases in soil strength. Water movement in soils

slows down because the pore size distribution shifts toward smaller pores. Smaller pores remain

saturated (water-logged) and thereby reduce gas exchange between the soil and the atmosphere,

which restricts the normal uptake of soil water by roots. High soil strength, especially as soil dries,

prevents root penetration to deeper layers to access stored soil water.

Classical calculations of plant available water ignore many restrictions imposed by soil factors. In

Chapter 2 of this thesis, I evaluated the Integral Water Capacity (IWC) model of Groenevelt et

al.(2001); (2004) using a set of undisturbed soil cores taken from the field. The evaluation included

applying weighting factors to attenuate the differential water capacity for osmotic stress, high soil

penetration resistance, poor soil aeration and declining soil hydraulic conductivity. The estimates of

plant-available water from this model are theoretically based and they rely on various assumptions

that have never been tested and may or may not hold in reality. In this Chapter I examine the extent

62

to which the estimated amounts of plant-available water match the total amount of water extracted

from the same soil by a relatively salt-tolerant plant, Rhodes grass (Chloris gayana cv. Pioneer).

I started with the research question: “Does the amount of plant-available water predicted from a set

of soil cores in the laboratory using the IWC-model of Groenevelt et al.(2001; 2004) match the

amount of water extracted by plants growing in situ in the same soil under field conditions?” The

corresponding null hypothesis to address this question was: “Extraction of water by plants under

field conditions matches that predicted to be ‘available’ using the IWC model”. The alternative

hypothesis is, of course, that the amount of water extracted by plants in the field exceeds or

underestimates that predicted by the IWC.

3.2 Materials and methods

3.2.1 Experimental design

In the same saline area where the soil samples were taken for the laboratory analyses (outlined in

Chapter 2) an area of 10 m x 20 m was cleared of its native grasses to establish three experimental

plots of 3.5 x 3.5 m each and separated by paths 0.5 m wide (Figure 3.1). All three plots were used

for calculation of plant available water in the soil profile but only the central plot was used to

calibrate the neutron probe by destructive sampling throughout the growing season. The central

Plot 2 was also used for root sampling.

Figure 3.1 Diagram of experimental plots showing dimensions and locations of neutron access tubes.

Plot 1 Plot 2 Plot 3

Paths

3.5 m

3.5

m

0.5 m 0.5 m

0.5 m

Isolation trenches lined with polyethylene to 1.5 m

Neutron access tubes

63

To obtain a water balance for the three plots, they were isolated from the surrounding area (and

from each other) by excavating trenches to 1.8 m deep on all sides, then lining each plot with thick

plastic sheeting before back-filling (Figure 3.2).

Figure 3.2 Preparation of the three isolated field plots for complete profile saturation and planting of Rhodes grass (Chloris gayana cv. Pioneer).

64

3.2.2 Water balance model

The model used to create a water balance consisted of the usual components to describe the change

in storage of soil water, &S, as follows:

&S = (I + R + G + Li) – (T + E + D + Lo) , [3.1]

where the inputs (mm) are I = irrigation, R = rainfall, G = groundwater accession, Li = lateral input

of water, and where the outputs (mm) are T = crop transpiration, E = evaporation from the soil

surface, D = subsoil drainage, and Lo = lateral output of water.

Apart from the initial irrigation, all inputs and outputs of water in Equation [3.1], except &S and T,

were assumed to be negligible on the following basis. Groundwater accession, G, was unlikely

because the regional water table was far deeper than the base of the plots (20 m). Lateral inputs and

outputs, Li and Lo, were prevented by the plastic sheeting that isolated each plot. Subsoil drainage,

D, was assumed to be prevented by the sodic, heavy clay subsoil horizon, which formed an

effective natural hydraulic barrier at about 1.5 m depth. The only water allowed onto each plot was

controlled by irrigation, I (see below); the plots were covered by tarpaulins to shed rain, R, when

required (see below). From the beginning of the critical experimental period, all soil surfaces were

completely covered with a dense crop, so soil evaporation, E, was considered to be negligible.

When irrigation ceased, each plot was therefore assumed to be ‘closed’ with respect to all inputs

and outputs of water except by transpiration, T, and the water balance was reduced to:

T = -&S [3.2]

For the initial irrigation, I, a watering system was established on each plot to bring the water

content of the entire profile up to saturation. Thirty metre lengths of 13 mm diameter drip irrigation

tubing (with drippers spaced 0.5 m apart) were placed on the soil surface from the centre of each

plot in an outward spiral (Figure 3.2). The end of each tube was crimped at the plot centre and the

other end was connected to a water tank (1.0 m x 1.0 m x 0.8 m) set up to supply water at a small

positive head (ca. +50 cm) by syphoning (Figure 3.3).

It was calculated that 8 m3 of water was required to completely saturate each experimental plot (3.5

m x 3.5 m) to a depth of 1.8 m (based upon total porosities calculated from the measured bulk

density and particle density of each horizon shown in Table 2.2. This volume of water was applied

to each plot over a period of two weeks during October 2010 after which the soil surface was

65

allowed to dry for a few days. Loose, soft seedbeds were then prepared using a hand rake, and 10g

lots of perennial Rhodes grass (Chloris gayana) were mixed with 500 g coarse sand and broadcast

onto each plot5. During germination and establishment, irrigation continued, thinning and

transplanting were performed to achieve uniform and full soil coverage and weeds were removed to

establish uniform coverage of Rhodes grass (Chloris gayana cv. Pioneer). Water was supplied to

keep the profile very wet to 1.8 m for several months until full canopy coverage occurred with a

leaf area index of at least 4.0 m2 leaves per m2 soil. When the canopy was full, light interception at

the base of the canopy was measured using an LP-80 light interception meter, which confirmed that

96% (standard deviation ± 1%) of solar radiation was being intercepted by the Rhodes grass

(Chloris gayana cv. Pioneer). At this point, one final irrigation was applied on 25 January 2011 and

no further water was allowed to enter the plots. After 48 hours (e.g. 27 January 2011) the first

neutron probe reading within the experimental period were taken.

Figure 3.3 Water supply system, rain-shelter frame, taking readings with neutron probe.

5 Five g of additional seed was germinated in trays to transplant into areas where plant establishment in the plots was variable or low.

66

When irrigation stopped (I = 0), the entire soil profile was considered to be at a nominal field

capacity, such that all subsequent changes in water content, &S, were considered to have occurred

by transpiration, T. A metal frame was set up over the plots to allow short-term coverage and

prevent any further input of water after irrigation ceased (R = 0). When rain was expected during

the experimental period after irrigation stopped, the frame was used to secure a tarpaulin to shed all

rain and then remove it during fine weather, which was most of the time (Figure 3.4).

Figure 3.4 Canvas suspended from rain shelter to shed any rain when expected (not often).

Storage of soil water, &S, was monitored by measuring the volumetric water content of the soil

regularly throughout the experiment. Five neutron access tubes (1.5 m long and 0.05 m diameter)

were installed in each of Plots 1 and 3 to a depth of 1.4 m, while only two access tubes were

installed in centre Plot 2 (see Figure 3.1). Plot 2 was used to collect undisturbed soil cores down the

67

profile at the same time each water content reading was taken using the neutron probe. The soil

cores were used to calibrate the neutron moisture meter (a Campbell Pacific Nuclear Model

CPN503 Hydroprobe) at this site. A 50 mm diameter PVC cap was placed over the top of each of

the 12 neutron access tubes to prevent water entry for the duration of the experiment. The water

content down the entire soil profile was then monitored once each week from this point for several

months throughout the summer and autumn period until all the plants died from water stress and

dehydration (Figure 3.5).

Figure 3.5 Photographs of the perennial Rhodes grass (Chloris gayana cv. Pioneer) plots from the last irrigation (27 Jan 2011) until the plants stopped extracting water and never recovered after rainfall (15 June 2011).

68

Neutron moisture readings were taken in each of the 12 tubes at 0, 5, 15, 25, 33, 68 and 139 days

after irrigation stopped. Five6 x 15 second neutron moisture readings were taken at 10 cm depth

intervals to 140 cm, starting with the bottom layer (140 cm depth) and then increasingly shallow

layers.

After the soil moisture readings were taken each day, two complete soil cores were taken from the

centre plot down to 1.4 m using a drilling rig. Replicate samples were extracted from the cores in

every horizon and taken to the laboratory to measure water content and bulk density to calibrate the

CPN 503 Hydroprobe (see section 3.4.4 below).

On the final sampling date, four additional soil cores (diameter 43 mm) were taken from the centre

Plot No. 2 to a depth of 150 cm to measure the mass of root tissue per unit volume down the

profile. The soil cores were sectioned into 20 cm increments7, sealed quantitatively in plastic bags

and transferred to a laboratory freezer set at -20oC for temporary storage. Twenty four hours before

analyses was planned, the samples were withdrawn from the freezer to thaw at room temperature.

Once thawed, samples were placed into square containers (2 litre volume) to soak in 1 litre of water

for 1 hour to allow fragmentation of the soil core and for roots to float to the surface for separation.

The larger (floating) roots were collected using tweezers and the remaining suspension of soil and

roots transferred to a nest of 20 cm diameter brass sieves (1 mm mesh over 0.5 mm mesh) to

capture smaller roots. A further 1 litre of water was used to continue the soil root separation and

passed through the nest of sieves and this was repeated until no further roots were collected on the

sieves. The roots were then washed off the sieves into 250 mL containers and placed in an oven set

at 70oC to dry for 4 days. The dry mass of roots was recorded and the mass-density calculated

based upon the volume of the 20 cm long core of soil from which the roots came (mg cm-3).

3.2.3 Calibrating the CPN 503 Hydroprobe neutron moisture meter.

Aside from protons in water, the slow neutron count rate is affected by the presence of other large

atoms such as boron, iron and chlorine (Marshall et al. 1996). Under the saline soil conditions used

in this study, the abundance of chloride and sometimes boron causes fast neutrons to be absorbed;

this reduces the slow neutron count and thereby the accuracy of the volumetric water contents. The

concentration of the important elements was measured on saturation paste extracts (Table 3.1),

which indicated that although there was some boron present it was not concentrated enough to

6 The number of replicate neutron counts was based upon a statistical analysis outlined in Section 3.4.4. 7 The purpose of this exercise was to assess the depth of root penetration, not to measure root density in each soil horizon. Hence samples were not taken at the same intervals as neutron probe reading or within horizons. This saved a considerable amount of work.

69

interfere significantly with the slow neutron count, and there was no detectible cadmium present.

The calibration therefore focussed solely on the presence of chlorine.

Table 3.1 Chemical properties (saturated paste extracts) of the soil profile in the plots containing the neutron access tubes.

Horizon (cm) pH EC, dS/m B, mg/L Fe, mg/L Ca Mg Na K

SAR mg/L mg/L mg/L mg/L

0 - 10 7.83 4.10 0.40 0.13 350 87 154 560 2

10 - 25 7.32 7.04 0.53 < 0.02 570 168 580 280 5

25 - 35 7.35 6.01 0.51 < 0.02 460 160 660 27 7

35 - 55 7.40 6.07 0.72 < 0.02 320 157 730 4.8 8

55 - 75 7.43 5.00 1.30 < 0.02 132 91 850 5.6 14

75 - 100 7.90 3.12 3.90 < 0.02 13 16 680 4.9 30

100 - 115 7.98 3.19 6.20 < 0.02 6.2 12 690 5.2 37

115 - 150 7.79 4.18 7.20 < 0.02 8.0 18 880 7.0 39

> 150 7.66 4.42 6.40 < 0.02 7.5 20 970 8.2 42

To allow correction of the slow neutron counts in the presence of chloride from salt, the standard

slow neutron count rate was determined in the laboratory using different concentrations of NaCl in

a 200 L drum of water. A 1.2 m neutron access tube was fitted inside the base of the drum and held

at the top by passing it through a tight hole in the lid. The drum was filled with RO water and

twenty slow neutron counts taken by lowering the neutron source down the access tube to a depth

of 0.5 m from the top of the barrel. Appropriate amounts of salt were progressively dissolved in the

water to make solutions of different salt concentration including 1, 2, 3, and 4 g/L of NaCl, and the

procedure repeated.

The standard count rates, CRs, were plotted as a function of the salt concentration converted to

electrical conductivity, EC (dS m-1) (Figure 3.6), which produced a negative linear correlation, as

expected:

CRs = -98.376 EC + 22819 (R² = 0.9911) , [3.3]

which shows that the standard slow neutron count rate, CRs, declined by approximately 100 slow

neutrons per unit increase in EC (i.e. 0.4%). This reduction is known as the chlorine effect (Wells

and Fityus 2011). Although the reduction in the standard rate is not large (approximately 3.5% over

the relevant range of salinity), if it is not taken into account it can be responsible for greater errors

in water content as the soil dries out. The value for the standard count rate, CRs, for each soil count

rate, CR, was therefore always adjusted using Equation [3.3] after the EC of soil samples were

70

collected each week during the drying period. The measured ECs were taken on 1:5 soil:water

extracts, so these were first corrected to the EC of paste extracts using Slavich and Petterson (1993)

method.

Figure 3.6 Mean standard 15 second count rate, CRs, of CPN 503 Hydroprobe in a large drum of water having different salt concentrations as measured by EC (dS m-1). The red vertical bars through each point represent the ± standard error of the mean of 20 readings.

The variability in slow neutron counts was not influenced by salt concentration. That is, there was

no trend in the measurement error with salt content. For example, the sample variance of the mean

neutron count rate, s2, ranged between 13806 and 22620, which produced very low standard errors

and coefficients of variation (Table 3.2).

Table 3.2 Variation in the standard 15-second slow-neutron count rate with salt concentration.

[NaCl] mg/L

Mean count rate, CRs

Standard deviation of mean CRs

Number of measurements

Coefficient variation of

mean CRs, %

Standard error of

mean CRs 0 22792 147.9 20 0.65 33.1

1000 22603 117.5 20 0.52 26.3 2000 22423 138.1 20 0.62 30.9 3000 22221 129.9 20 0.58 29.0 4000 22030 150.4 20 0.68 33.6

By choosing an ‘acceptable’ error in slow neutron count readings, it was possible to calculate the

number of readings, n, required to obtain acceptably accurate mean neutron counts for a given

instrument, using the relation (Snedecor and Cochran 1989):

, [3.4]

71

where s was the sample variance, and L was the upper limit to the desired amount of error tolerated

in a given neutron reading. Obviously the larger the tolerance for error the smaller the number of

readings required. Based on the data reported in Table 3.2, Equation [3.4] suggested that 6 to 9

readings would be required to ensure the error in count rates stayed within 100 slow neutrons,

while only 2 to 4 readings would be required to ensure the error in count rates stayed within 150.

My experience in the field based on count rate suggested that an error of between 100 and 150

counts would be quite acceptable, so a compromise was chosen to be 5 readings at each soil depth.

To calibrate the CPN 503 probe, all 5 readings of the soil count rate, CR, taken in the field at each

depth were averaged and then divided by the standard count rate in the appropriate salt solution

according to Equation [3.3], CRs, to produce a relative count rate, RCR, according to the relation:

[3.5]

For every value of RCR there was a corresponding value of the true volumetric water content of the

soil in the field, �. For each horizon, there was thus a complete set of RCR and � values for the soil

taken over the growing season from the wet state on 27 January 2011 down to the wilting point on

15 June 2011. These points were plotted to produce a set of either quadratic or cubic regression

equations between RCR and water content measured in the laboratory for each soil horizon (Table

3.3). As can be seen from the very high regression coefficients, the correlations between the

relative count rate and the measured volumetric water content is excellent for all the measurements

(R2 > 0.95 with one exception). The least excellent correlation occurred for the readings taken in

the top 10 cm, where the correlation coefficient was only R2 = 0.735 (Table 3.3). The poorer

correlation for the surface horizon was, of course, expected and much has been written on how to

overcome the problem of lost neutrons – manufacturers always recommend using neutron moisture

meters only for depths > 20cm below the soil surface. Because corrections and other means of

avoiding neutron losses are not particularly simple, I decided to take additional (destructive)

samples from the top 10 cm for this experiment, using small cores (5 cm x 5 cm) rather than relying

on corrections.

3.3 Results and discussion

3.3.1 Plant water use from full canopy establishment to plant death

The progression of plant vigour for the Rhodes grass (Chloris gayana cv. Pioneer) stand from its

fully established canopy to its death brought on by water stress is shown in Figure 3.5. The

corresponding soil water extraction patterns are shown in Figure 3.7 for the three field

experimental plots. The water content profiles for Plot 1 are compared with those of the central Plot

72

2 in Figure 3.7a; the profiles for Plot 3 are also compared with those of the central Plot 2 in Figure

3.7b. Finally, because all three plots showed essentially the same patterns over time, all three

profiles were averaged to produce a single set of temporal water content profiles in Figure 3.7c,

around which discussion will be focussed.

Table 3.3. Correlations between relative slow neutron count rate, RCR, and volumetric water content, �, at each depth in the soil profile.

Depth, (cm)

Correlation between volumetric water content and relative count rate from the neutron probe

Regression coefficient, R2

10 � = -9.68 RCR2 + 2.65RCR 0.735 20 � = 2.49 RCR3 - 5.44 RCR2 + 3.99 RCR - 0.61 0.935 30 � = 26.49 RCR3 - 34.69 RCR2 + 15.50 RCR – 2.13 0.993 40 � = 42.02 RCR3 - 60.29 RCR2 + 29.18 RCR – 4.50 0.993 50 � = 20.91 RCR3 - 28.89 RCR2 + 13.64 RCR – 1.94 0.983 60 � = 6.03 RCR3 - 8.14 RCR2 + 4.21 RCR – 0.55 0.992 70 � = 5.15 RCR3 - 6.84 RCR2 + 3.60 RCR – 0.44 0.997 80 � = 17.44 RCR3 - 22.46 RCR2 + 9.84 RCR – 1.22 0.991 90 � = 12.48 RCR3 - 16.43 RCR2 + 7.55 RCR – 0.94 0.996 100 � = 23.64 RCR3 - 31.11 RCR2 + 13.76 RCR – 1.79 0.964 110 � = 27.03 RCR3 - 36.22 RCR2 + 16.37 RCR – 2.22 0.952 120 � = 6.86 RCR3 - 8.91 RCR2 + 4.24 RCR – 0.44 0.996 130 � = 12.74 RCR3 - 17.62 RCR2 + 8.48 RCR – 1.11 0.962 140 � = 13.43 RCR3 - 21.01 RCR2 + 11.33 RCR – 1.78 0.980

3.3.2 Plant water use and root distribution

Assuming Equation [3.2] is valid and that all changes in the volumetric water content down the

profiles resulted from Rhodes grass transpiration alone, the greatest amount of active root water-

extraction occurred in the top 100 cm. The relatively slow reductions in water content below 100

cm in Figure 3.7c suggest either that roots were not very dense from that depth downward, or that

they were restricted in some way and were unable to extract as much water as the roots did in the

upper layers.

Root mass density measurements confirmed that roots were present all the way down the profile

(Figure 3.8) albeit at lower densities below 100 cm (Sheldon and Menzies 2005), and that these

roots did not extract as much water. Changes in the water content in the 100 to 140 cm zone may

simply reflect some upward capillary rise to replenish water use in the active part of the root zone,

but a complete evaluation of this idea is not within the scope of this thesis.

73

Figure 3.7a Volumetric water content as a function of depth for plots 1 and 2, from the time of the initial profile saturation (03 Nov 2010) until the plants wilted completely (15 June 2011). Horizontal bars represent ± standard error of the mean water content.

0

20

40

60

80

100

120

140

0.00 0.10 0.20 0.30 0.40 0.50

Dep

th, c

m

�v, cm3/cm3

0

20

40

60

80

100

120

140

0.00 0.10 0.20 0.30 0.40 0.50D

epth

, cm

�v, cm3/cm3

�v at saturation; these values are not included in the calculation of PAW


Plot 1

Plot 2 (centre)

74

0

20

40

60

80

100

120

140

0.00 0.10 0.20 0.30 0.40 0.50D

epth

, cm

�v, cm3/cm3

Figure 3.7b Volumetric water content as a function of depth for plots 3 and 2, from the time of the initial profile saturation (03 Nov 2010) until the plants wilted completely (15 June 2011). Horizontal bars represent ± standard error of the mean water content.

0

20

40

60

80

100

120

140

0.00 0.10 0.20 0.30 0.40 0.50

Dep

th, c

m

�v, cm3/cm3

Plot 3

Plot 2 (centre)



75

Figure 3.7c Volumetric water content as a function of depth for plots all three plots averaged, from the time of the initial profile saturation (03 Nov 2010) until the plants wilted completely (15 June 2011). Horizontal bars represent ± standard error of the mean water content.

Figure 3.8 Distribution of Rhodes grass root mass per unit volume as a function of depth below the soil surface.

0

20

40

60

80

100

120

140

0.00 0.10 0.20 0.30 0.40 0.50D

epth

, cm

�v, cm3/cm3

All 3 plots averaged


76

3.3.3 Evaluation of the IWC model against water use by real plants

If one considers the maximum reduction in water content over the growing season to represent the

total amount of plant-available water held in the soil for a given crop, this amount of water can be

compared to the predictions produced by the IWC model presented in Chapter 2. The difference in

volumetric water content down the soil profile from when the soil was at field capacity with a fully

developed canopy and root system (27 January 2011) and the time when the plants wilted (11 June

2011) is given in Table 3.4 along with estimates of the IWC at each depth as well as the classical

estimate of PAW. Comparisons are made easier by superimposing the field-measured extraction

data onto the predictions made in Chapter 2 (now Figure 3.9), which clearly shows that the PAW

estimate, the simplest of all models having no attenuation of the water capacity, gave the closest

estimate of the real water extraction by Rhodes grass in this saline soil. This finding may seem

surprising at first but Rhodes grass is a relatively salt tolerant plant, which means that weighting

the water capacity for salt stress is unwarranted. As seen in Chapter 2, the weighting for salt is the

most severe of any attenuation so if the analysis is repeated using the differential water capacity

instead of the soil solution capacity, the attenuations are more modest (Table 3.5; Figure 3.10).

Table 3.4. Predictions of plant available water in a saline soil profile (mm/m) based upon various degrees of weighting of the differential water capacity (taken directly from Table 2.3) compared with field-measured change in water contents with Rhodes grass.

Depth cm

Field measurements Predicted water availability

��max �min &� PAW

No attenuation

IWC Salt

alone

IWC Salt +

Aeration

IWC Salt +

SR

IWC Salt + K(hm)

IWC All

factors mm / m

0-5 390 70 320 284 124 109 98 120 81 5-10 420 210 210 284 124 109 98 120 81 10-20 310 160 150 132 84 8 48 74 3 20-30 330 180 150 137 84 22 57 78 12 30-40 350 180 170 164 93 47 63 89 26 40-50 340 170 170 187 101 59 61 95 30 50-60 320 140 180 204 108 72 61 105 33 60-70 340 140 200 221 114 84 61 115 36 70-80 320 140 180 237 111 84 61 103 35 80-90 320 150 170 252 108 83 61 90 34

90-100 310 160 150 252 108 83 61 90 34 100-110 320 170 150 128 89 18 51 78 5 110-120 330 230 100 103 73 11 41 66 3 120-130 340 250 90 78 57 5 30 53 0 130-140 370 330 40 78 57 5 30 53 0

77

Table 3.5. Predictions of plant available water (mm/m) based upon the same weightings of the differential water capacity but ignoring salt, compared with field-measured change in water contents with Rhodes grass.

Depth cm

Field measurements Predicted water availability

��max �min &� PAW

No attenuation

IWC Aeration

only

IWC SR only

IWC K(hm) only

IWC All factors

mm / m 0-5 390 70 320 284 189 243 270 151 5-10 420 210 210 284 189 243 270 151 10-20 310 160 150 132 12 75 114 3 20-30 330 180 150 137 44 116 137 28 30-40 350 180 170 164 80 121 176 39 40-50 340 170 170 187 80 121 176 39 50-60 320 140 180 204 80 121 176 39 60-70 340 140 200 221 117 124 210 51 70-80 320 140 180 237 117 124 210 51 80-90 320 150 170 252 150 160 212 61 90-100 310 160 150 252 150 160 212 61 100-110 320 170 150 128 22 79 112 6 110-120 330 230 100 103 22 79 112 6 120-130 340 250 90 78 1 43 72 0 130-140 370 330 40 78 0 52 69 0

Figure 3.9. Real soil water extraction for Rhodes grass (Chloris gayana cv. Pioneer) superimposed on estimates of water availability after various weightings of the soil solution capacity (i.e. including consideration of salt).

PAW – no factors Salt Salt + K(h) Salt + soil resistance Salt + aeration Salt + all factors Rhodes grass

78

Figure 3.10. Real soil water extraction for Rhodes grass (Chloris gayana cv. Pioneer) superimposed on estimates of water availability after various weightings of the soil water capacity (i.e. excluding consideration of salt).

Again, however, it becomes clear that even when osmotic stresses are ignored, the IWC approach

to attenuating the water capacity generally predicts significantly less water should be available than

the Rhodes grass actually extracted. There are several possible explanations for this. The most

obvious explanation is that the coefficients selected in the weighting functions were too severe in

this analysis. With absolutely no knowledge of the ‘real’ coefficients, the middle coefficient (of the

3 used in demonstration) was chosen. If the least severe coefficient of the 3 is chosen, all the lines

move closer to the ‘real’ data (Figure 3.11).

Even so, the correction shown in Figure 3.11 is only modest, which suggests the order of

magnitude of the coefficients in the weighting functions may not be quite right. By shifting the

coefficients by a full order of magnitude, most of the individual estimates moved significantly

closer to the PAW estimate and somewhat closer to the ‘real’ data (Figure 3.12). Without new

experimental data for different crops and soils, however, further manipulation of the coefficients in

the weighting functions is unlikely to yield much progress.

PAW – no factors K(h) only Soil resistance only Aeration only All factors except salt Rhodes grass

79

Figure 3.11. Comparative estimates of water availability from Figure 3.10 adjusted with ‘gentler’ coefficients in the weighting functions.

Figure 3.12. Comparative estimates of water availability from Figure 3.11 adjusted with significantly ‘gentler’ coefficients in the weighting functions.



80

3.4 Conclusions

Even though the soil profile at this site was saline, estimates of plant-available water using IWC

with a weighting function for salinity proved too severe for the relatively salt tolerant Rhodes grass

(Chloris gayana cv. Pioneer). Adjusting the coefficients �, A and ' in the weighting functions to

minimize the attenuating effects of high soil resistance, low aeration and low hydraulic

conductivity, respectively, was only partly successful in bringing the IWC into line with measured

plant behaviour.

To make a thorough evaluation of the utility of weighting the water capacity for salinity with the

IWC, it will be necessary to use a range of different plants that vary in their tolerance of salt. Use

of very salt-sensitive plants in the first instance would go some way toward evaluating the merits of

the IWC model. However, this is easier said than done. The difficulty with such an approach would

be that to evaluate how much water a salt-sensitive root system can extract from a saline soil, a root

system must first be established in a medium in which it would not normally grow! A potential

solution might be to grow salt-sensitive plants in a non-saline medium and then expose it to salt

once the root system is fully established.

It is also possible that the assumptions used to establish some of the limits of integration in the

IWC model (e.g. hm = 10 kPa at the wet end, and hm = 1500 kPa at the dry end) are flawed. This

possibility needs to be evaluated using plants to establish the real limits. For example, given that

the water capacity has the greatest magnitude at the wet end, it is possible that plants can extract a

great deal more water before it drains away than is currently thought. Similarly, at the dry end,

some species are able to extract soil water well past the point that is nominally considered to be the

permanent wilting point. In this study, Rhodes grass (Chloris gayana cv. Pioneer) appeared to take

the soil matric head well past the point predicted from the water retention curve at 1500 kPa.

Greater understanding of the behaviour of a range of plants at the dry end would go some way

toward establishing plant-specific limits of integration.

81

Chapter 4 Changes in IWC during reclamation of a salt-affected soil

4.1 Introduction

Salt-affected soils are among the least productive soils in the world because they are often saline,

sodic and/or alkaline, which impedes plant water and nutrient uptake from them. Salinity, sodicity

and alkalinity generate osmotic stresses, element toxicities and poor soil physical conditions. In

particular, water may move slowly through them, be held in small (anaerobic) pores and the soil

may become hard and impenetrable to roots (Shainberg and Shalhevet 1984). To become

agriculturally productive, salt-affected soils must be reclaimed by draining and leaching them, and

by increasing the quantities of organic matter and calcium throughout the soil profile (e.g. (Jury et

al. 1979; Qadir et al. 1996; Ilyas et al. 1997; Akhter et al. 2004)

Reclamation of salt-affected soils, however, is no simple matter because any reduction in salt

concentration through leaching without a concomitant addition of divalent cations (e.g. calcium) to

replace sodium generally results in an immediate deterioration of soil hydraulic properties (c.f.

Chapter 1). Soils take on a massive structure with greatly reduced infiltration and hydraulic

conductivity and eventually become uninhabitable by plants. However, if dilute calcium salts are

gradually added such that the hydraulic conductivity is not reduced significantly during drainage

and leaching, the sodicity can be reduced at the same time that salt is removed and the physical

properties of the soil can gradually be improved (Quirk and Schofield 1955; Quirk 1986; Gupta and

Singh 1988; Rengasamy and Olsson 1991)

The improvement in soil properties during reclamation can dictate the productive capacity of the

soil for plants, especially on heavy clay-textured soils (Blokhuis 1980). The gradual success of

plants on reclaimed soils is due to a combination of improvements in nutrient availability, as well

as chemical and physical properties. In most cases, the re-introduction of plants occurs after

disturbance of the soil by some sort of tillage, the structure of which becomes stabilized to a large

extent by calcium and new plant roots. However, the extent to which improvements in soil

properties can be achieved without any tillage or other soil disturbance (i.e. simply by leaching

with appropriate salt solutions) is important to understand where deep tillage and disturbance are

neither feasible nor desirable.

If it is possible to make spontaneous changes in pore size distribution and structure simply by

leaching with calcium solutions, the magnitude of such changes should be reflected in the water

retention curve and possibly in other relevant soil physical properties. Jayawardane and Beattie

(1979) showed that leaching with solutions of differing SAR and EC made significant changes to

82

the water retention curve, but their study focussed solely on pore size distributions and ignored all

other soil physical properties and made no mention of plant available water. In the present study I

asked “how does soil water availability change (as measured using the IWC) during various stages

of reclamation of a saline-sodic soil?” I hypothesized that IWC would increase monotonously from

one critical state to the next during reclamation, but I acknowledged that an alternative hypothesis

would be that IWC would increase non-monotonously, depending upon the relative importance of

various soil physical properties at each stage of reclamation.


4.2.1 Experimental approach and design

This set of laboratory experiments was designed to take soil of the 9 horizons used in Chapters 2

and 3 from their initial saline-sodic states to various different, less saline and less sodic states, then

to measure the consequent changes in soil physical properties and determine whether they

influenced the IWC significantly.

Six different pathways (treatments) were chosen to take the soils from the saline-sodic state to a

lesser saline-sodic state. These consisted of a preliminary treatment applied to all samples of a

given horizon followed by a further treatment to reduce the salinity or sodicity:

1) Initial soil EC: Leaching with a solution having an EC & SAR to match that in the field; no

further leaching with any other solution – this was the ‘control’ treatment.

2) RO water: Treatment 1 followed by leaching with RO water – this was to maximize swelling

and dispersion and produce the most extreme effects on soil physical properties.

3) 0.1 M CaCl2: Treatment 1 followed by leaching first with 0.1 M CaCl2, then RO water – this

was to generate the extreme opposite effect to Treatment 2.

4) 1/2 initial EC: Treatment 1 followed by leaching first with a solution of identical composition

to Pre-treatment 1 but with only one half the initial EC (and ) this was to generate one

of three intermediate effects between Treatments 1 and 2.

5) 1/4 initial EC: Treatment 1 followed by Treatment 4 followed by leaching with a solution of

only one quarter the initial EC (and ) – this was to generate the second of three

intermediate effects between Treatments 1 and 2.

6) 1/16 initial EC: Treatment 1 followed Treatment 5 followed by leaching with a solution of

only one eighth the initial EC (and ) , followed by leaching with a solution of only

one sixteenth the initial EC (and ) – this was to generate the third of three intermediate

effects between Treatments 1 and 2.

83

The experimental design comprised 9 soil horizons x 6 treatments per horizon in a completely

randomized design with 4 replicate-soil cores (sub-units) buried within a container (pot) of soil for

each treatment (explained below).

4.2.2 Experimental units

The experimental sub-units consisted of 4 x small cylinders of soil (50 mm x 50 mm) buried within

each ‘pot’ of soil. There was one pot per horizon for each treatment, yielding a total number of 54

pots. The pots were constructed using cylindrical PVC drainage piping (internal diameter = 152.5

mm, height = 200 mm). A cylindrical base-cap having multiple 3-mm-diameter holes drilled to

allow leachate collection was connected to one end of each pot and the base was lined with plastic

screening plus Whatman-42 filter paper to prevent soil loss through the drain holes.

The soil from each of the nine soil horizons was air dried, passed through a 5 mm sieve then stored

in 20-litre plastic buckets. After some preliminary work, the equivalent dry mass of soil applicable

to the maximum density each soil horizon could reasonably be taken up to was packed into the pots

as follows. Approximately 30% of the required soil mass was packed into the bottom third of each

pot and then 4 x stainless steel cylindrical rings (inside diameter = 50 mm; height = 50 mm) were

placed upright on top of this soil in each pot. The remaining soil was then packed in and around the

4 rings such that they were embedded in each pot with 50 mm of soil above and below them

(Figure 4.1). The final bulk densities achieved in the pots are shown in Table 4.1. these bulk

density were necessarily less than those in the field (Table 2.2) by 5-15%. However, this

experiment was designed to observe relative changes in soil physical properties during reclamation

that might influence IWC

Figure 4.1 Dimensions of experimental pot of soil with 4 small soil cores embedded.

84

Table 4.1. Bulk densities achieved for 150 mm columns of soil from each soil horizon (cylindrical pot diameter = 152.5 mm; calculated volume of each soil column = 2740 cm3).

Horizon, cm

Mass of soil, g

Bulk density, g/cm3

0 - 10 2603 0.95 10 - 25 3671 1.34 25 - 35 3973 1.45 35 - 55 4000 1.46 55- 75 4000 1.46

75 - 100 4055 1.48 100 - 115 4082 1.49 115 - 150 3973 1.45

> 150 4000 1.46

4.2.3 Experimental protocol

The appropriate composition of the leaching solutions was determined from an analysis of each soil

horizon. In this study (as distinct from the study reported in Chapter 2) all analyses were performed

on saturation paste extracts, the preparation of which is described in Appendix 18. Field-isotonic

solutions were prepared based upon the SAR and total cation concentrations shown in Table 4.2.

Table 4.2 Elemental analysis by ICP-MS for the major cations, plus SAR, EC and pH of the saturation paste extracts in each of the 9 soil horizons. SAR was calculated by dividing [Na] (mmol/L) by the square root of ([Ca] + [Mg]). The value for � cations (mmolc L-1) was the sum of ([Na] + [K]) plus twice the sum of ([Ca] + [Mg]). The values of ECmeas were measured and they compare well with the values for ECcalc, which were calculated from � cations divided by 10.

Soil depth Ca Mg Na K SAR �� cations ECcalc ECmeas pH cm mmol L-1 mmolc L-1 dS m-1

0 - 10 8.75 3.64 6.71 14.36 2 45.85 4.59 4.10 7.83 10 - 25 14.25 7.01 25.00 7.18 5 74.70 7.47 7.04 7.32 25 - 35 11.25 6.55 28.26 0.69 7 64.55 6.45 6.01 7.35 35 - 55 8.25 6.80 32.61 0.13 8 62.84 6.28 6.07 7.40 55 - 75 3.31 3.84 37.17 0.15 14 51.62 5.16 5.00 7.43 75 - 100 0.33 0.67 29.78 0.13 30 31.91 3.19 3.12 7.90 100 - 115 0.16 0.51 30.43 0.15 37 31.93 3.19 3.19 7.98 115 - 150 0.22 0.79 39.57 0.19 39 41.77 4.18 4.18 7.79 > 150 0.19 0.83 42.39 0.24 42 44.68 4.47 4.42 7.66

Each of the 54 pots was wetted with the appropriate isotonic solution (from Table 4.2) by capillary

action established in large rectangular baths containing a few centimetres of each solution. When

8 The study reported in Chapter 2 used 1:5 soil:water extracts to measure pH and EC and then converted to equivalent saturation paste-extract values using the method of Slavich and Petterson (1989); this was done so that results could be compared across Chapters.

85

the soil surface of each pot became wet by capillary rise, the level of the solution in each bath was

gradually raised to the top of each pot to completely saturate the soil. When saturation was

achieved, Whatman-42 filter paper was placed on the surface of each soil and a ponded head of 20

mm was established using a Mariotte bottle over each soil surface to begin the leaching process.

Each pot was suspended over a container in which the leachate was collected to measure the EC

until it reached the same value as the input-solution EC and until the flux of solution reached

steady state (this required approximately 2 days – see below). At this point the saturated hydraulic

conductivity was determined from the measured flux of leachate, the cross-sectional area of the pot,

and the hydraulic gradient, determined from the height of the soil column and of the static pond of

solution on the soil surface. After the saturated hydraulic conductivity was measured, one of the 6

pots for each soil horizon was removed and allowed to stand and drain under the influence of

gravity for 24 hours (total of 9 pots) prior to taking further measurements (Figure 4.2).

Figure 4.2 Leaching and sampling protocol for each soil horizon. Treatment numbers are indicated in the first pot on the left.

6

5

4

3

2

1

86

Each of the 9 selected pots was then carefully dismantled to release the four embedded soil cores

(sub-units) for analysis as follows. The soil surrounding each of the four embedded cores in each

pot was trimmed away and a nylon mesh fitted to one end of each core held in a ring. These four

sub-units were then re-saturated in their respective isotonic solutions and then taken through the

steps required to obtain water retention curves and soil resistance curves as described in Chapter 2.

After the first set of pots was removed for analysis, the remaining 5 pots for each soil horizon were

taken through the next leaching stage according to the treatments listed in Section 4.2.1 above.

After each leaching-treatment, a complete set of 9 pots was removed and analysed as described

above to produce another set of water retention curves and soil resistance curves. This procedure

was continued until all 6 treatments were applied to all 9 horizons (replicated on 4 sub-units/cores)

to generate 54 x 4 = 216 soil water retention curves and soil resistance curves.

An estimate of the period required for cation exchange and diffusion to occur during leaching was

estimated from the approximate relation:

[4.1]

where is the root mean square distance travelled in a time t by diffusing cations in the soil

matrix where the effective diffusion coefficient is D. In the saturated soil matrices used in these

leaching experiments, D was estimated to be at least 3.3x10-11

m2 s-1 for the major cations (Rowell

et al. 1967; Gregory 2006). The diffusion path length was assumed to be equal to the radius of the

largest aggregates, which in this case was determined by the 5 mm sieve used in preparing the soil

(i.e. 0.0025 m). Equation [4.1] produces a time of 26 hours. It was decided to allow the soil

columns to leach for 48 hours before hydraulic conductivities were measured. At this time steady-

state flow rates were achieved.

The soil water retention and soil resistance data were fitted to the models of Groenevelt et al.

(2001; 2004) and Grant et al. (2010) using the mathematical software package, MathCAD 14

(Mathsoft 2008), and the integral water capacities (IWC) calculated.


4.3.1 Changes in saturated hydraulic conductivity during reclamation

As might be expected in a soil where the texture gradually becomes heavier with depth, the

saturated hydraulic conductivity, Ks, of the “Initial EC” (Treatment 1) soils, decreased by several

87

orders of magnitude down the soil profile from approximately 2 x 10-6 m s-1 in the top horizon

down to 5 x 10-8 m s-1 in the heavy clay subsoil (Figure 4.3). In every horizon down the profile, the

“Initial EC” soil (red bars, Treatment 1) and the “0.1M CaCl2” soils (pink bars, Treatment 3) had

the highest salt concentrations so they maintained the largest hydraulic conductivities. By contrast,

the “RO water” soils (blue bars, Treatment 2) had the lowest salt (electrolyte) concentrations, so

their hydraulic conductivities were invariably the lowest in all soil horizons. The differences in soil

solution among the intermediate treatments (dark grey bars, Treatment 4; light grey bars, Treatment

5; and white bars, Treatment 6) were reflected in the saturated hydraulic conductivities, which fell

between the three extreme treatments, although there were no consistent changes among

Treatments 4, 5 and 6 (Figure 4.3). These observations are all consistent with predictions arising

from theory and practice (e.g. Quirk and Schofield 1955); furthermore the trends were enhanced in

the sub-soil layers with greater clay contents, where a greater degree of swelling and dispersion of

clay and pore blockage due to deposition at points of particle contact (Dikinya et al. 2008) was

possible – it should be noted that the scale for Ks in the three sub soil horizons is an order of

magnitude smaller than in the upper horizons.

Figure 4.3 Changes in saturated hydraulic conductivity of repacked soil from a profile using leaching solutions of different EC and SAR.

88

4.3.2 Changes in water retention curves during reclamation

The water retention data from the 4 cores per pot were averaged and plotted for the 6 treatments

and 9 soil horizons in 54 water retention curves, each shown in Appendix 2. The water retention

model of Groenevelt et al. (2004) and Grant et al. (2010) was fitted to all the data and the

parameters are listed in Table 4.3. Where treatment effects were not obviously different, the data

were grouped and averaged to plot them in Figures 4.4a) to i) for the 9 soil horizons. For 8 of the 9

soil horizons, there was a clear general trend in the water contents at a given matric potential viz.

(Initial soil EC, 0.1M CaCl2) < ( ½, ¼, 1/16 initial EC) < RO water). Within these 3 treatment

groups, there were no significant differences. Accordingly, the curves have been grouped in this

way to exhibit extremes of behaviour and avoid clutter. This trend was absent from the 0-10 cm

horizon, probably because of its substantially larger soil organic matter content and coarser texture.

As found by Jayawardane and Beattie (1979), all treatments that generated significant swelling and

dispersion of clay (and caused the pore size distribution to shift toward smaller pores) also

increased water retention and produced an upward shift in the water retention curves. For example,

the soils treated with electrolyte-rich sodium solution alone (initial soil EC Treatment 1) or else 0.1

M calcium chloride (Treatment 3), remained completely stable and generated water retention

curves that declined more sharply (bottom red lines in Figures 4.4). This is consistent with the

finding that these soils all had significantly greater saturated hydraulic conductivities, which

reflected a distribution of larger, more stable pores (Figure 4.3). By contrast, the soils treated with

electrolyte-rich sodium solutions then diluted with high SAR solutions of lower salt content

(Treatments 2, 4, 5 and 6), all became unstable to varying degrees and produced finer pore size

distributions and less abrupt water retention curves. Again, this is consistent with the significantly

lower saturated hydraulic conductivities shown in Figure 4.3 for these soils.

89

Figure 4.4 Summary of water retention curves grouped according to whether treatment effects were obvious for soil horizons: a) 0-10 cm, b) 10-25 cm, c) 25-35 cm, and d) 35-55 cm. Groupings of curves are indicated for each soil horizon.

a) b)

c) d)

90

Figure 4.4 Water retention curves grouped according to whether treatment effects were obvious for soil horizons: e) 55-75 cm, f) 75-100 cm, g) 100-115 cm, and h) 115-150 cm. Groupings of curves are indicated for each soil horizon.

e) f)

g) h)

91

Figure 4.4 Water retention curves grouped according to whether treatment effects were obvious for soil horizon: i) > 150 cm. Groupings of curves are indicated.

The effects shown in Figure 4.3 and 4.4 depended on soil texture, as might be expected. For

example, the surface soil horizon, which contained mainly sand and organic matter, displayed

absolutely no significant treatment effects whatsoever, which is why there is only one (average)

water retention curve shown in Figure 4.4a. As the clay content increased with depth in the soil

profile (and as the organic matter content decreased), the treatment effects were gradually

magnified and the water retention curves separated from one another more obviously. Furthermore,

the variability in the water retention data dropped significantly as the clay content increased such

that the standard deviation of the mean �v at each suction was hidden to a large extent by the points

themselves (2nd and 3rd decimal places).

i)

92

Table 4.3 Fitting parameters for the Groenevelt et al. (2001) (2004) soil water retention curves shown in Figures 4.4, plus the penetration resistance curves for the 6 different leaching treatments in the 9 soil horizons shown in Figures 4.5.

Soil depth cm Treatment

Fitting parameters Water retention curve Penetration resistance

�s k0 k1 n a b 0 - 10 Initial EC (7.56 dS/m) 0.557 21.06 0.420 0.539 0.108 0.287

RO water 0.548 37.41 0.341 0.676 0.073 0.323 0.1M CaCl2 0.549 90.01 0.371 0.817 0.092 0.304 1/2 initial EC 0.539 24.87 0.395 0.520 0.089 0.303 1/4 initial EC 0.547 25.44 0.436 0.523 0.071 0.327 1/16 initial EC 0.560 7.00 0.477 0.347 0.065 0.336

10 - 25 Initial EC (12.29 dS/m) 0.470 68.50 0.244 0.898 0.229 0.264 RO water 0.479 27.01 0.188 0.635 0.285 0.224 0.1M CaCl2 0.474 34.70 0.246 0.777 0.174 0.293 1/2 initial EC 0.488 45.69 0.238 0.516 0.275 0.261 1/4 initial EC 0.478 101.49 0.246 0.829 0.199 0.29 1/16 initial EC 0.479 107.48 0.249 0.800 0.173 0.287

25 - 35 Initial EC (13.90 dS/m) 0.455 35.13 0.272 0.771 0.289 0.257 RO water 0.446 28.50 0.206 0.663 0.311 0.231 0.1M CaCl2 0.453 33.84 0.274 0.768 0.308 0.253 1/2 Initial EC 0.448 26.40 0.245 0.687 0.242 0.271 1/4 Initial EC 0.448 16.93 0.246 0.586 0.293 0.243 1/16 Initial EC 0.451 8.00 0.281 0.404 0.277 0.253






> 150 Initial EC (8.98 dS/m) 0.501 13.20 0.238 0.516 0.042 0.483 RO water 0.496 52.61 0.131 0.643 0.077 0.373 0.1M CaCl2 0.498 8.56 0.290 0.371 0.035 0.51 1/2 initial EC 0.498 30.23 0.196 0.600 0.082 0.398 1/4 initial EC 0.499 20.71 0.220 0.486 0.051 0.433 1/16 initial EC 0.494 22.73 0.224 0.466 0.037 0.464

93

4.3.3 Changes in soil penetration resistance during reclamation

The penetration resistance data from the 4 soil cores (sub-units) per pot were averaged and plotted

for the 6 treatments and 9 soil horizons in 54 soil resistance curves, all of which are shown in

Appendix 3. The power model describing the relationship between soil matric head and penetration

resistance (Equation [2.1]) was fitted to the data (they all followed a consistent shape), and the

parameters are listed in Table 4.3. Where treatment effects were not obviously different, the data

were grouped and averaged, and these are shown in Figures 4.5a) to i) for the 9 soil horizons. For 7

of the 9 soil horizons, there was a clear general trend in the soil penetration resistance at a given

matric potential viz. (Initial soil EC, 0.1M CaCl2) > ( ½, ¼, 1/16 initial EC) > RO water). Within

these 3 treatment groups, there were no significant differences. Accordingly, the curves have been

grouped in this way to exhibit extremes of behaviour and avoid clutter. This trend was absent from

the 0-10 cm and 10-25 cm horizons, probably because of the substantially larger soil organic matter

content of the former and the lower bulk densities of both horizons.

In general, the penetration resistance curves shifted upward with increasing clay content and

decreasing organic matter content (i.e. with depth in the soil profile). Importantly, the treatments

that caused greater swelling, dispersion and thus greater water retention, consistently reduced the

penetration resistance of the soil at any given matric head. That is, the treatments that flattened the

water retention curves (e.g. Treatments 2, 4, 5 and 6) also made the soil wetter and thus softer

(easier to penetrate). By contrast, the treatments that stabilized the soil and allowed it to drain under

suction, made it drier and thus harder to penetrate. This may seem counter-intuitive at first, because

it is usual to expect that as soil structure is degraded by excessive swelling and dispersion, it

becomes denser and harder to penetrate. However, wetting and drying cycles are generally required

to enhance the expression of such effects in nature, whereas the experimental protocol used here

did not allow for this.

Treatment effects were not detected in the top two horizons. In fact, there was little difference in

any of the treatment effects down to 100 cm, below which the differences gradually became

magnified.

94

Figure 4.5 Soil resistance curves grouped according to whether treatment effects were obvious for soil horizon: a) 0-10 cm, b) 10-25 cm, c) 25-35 cm, and d) 35-55 cm. Groupings of curves are indicated for each soil horizon.

b) a)

c) d)

95

Figure 4.5 Soil resistance curves grouped according to whether treatment effects were obvious for soil horizon: e) 55-75 cm, f) 75-100 cm, g) 100-115 cm, and h) 115-135 cm. Groupings of curves are indicated for each soil horizon.

e) f)

g) h)

96

Figure 4.5 Soil resistance curves grouped according to whether treatment effects were obvious for soil horizon: i) > 150 cm. Groupings of curves are indicated.

4.3.4 Changes in IWC during reclamation

All water retention curves were differentiated to obtain the water capacities, which were then either

integrated to produce the classical PAW, or weighted for various limiting soil factors and then

integrated to produce different measures of the IWC. Table 4.4 presents the values of PAW and

IWC for the 6 leaching treatments and 9 soil horizons. The IWC values are reported for each

limiting factor on its own as well as in combination with all other factors. Where the water

retention curves were not influenced by the treatments (see Figures 4.4), the relevant data were

averaged to prepare single water retention curves, and this is indicated by shading in Table 4.4.

Treatments that caused the water retention curve to become ‘flatter’ produced differential water

capacities that were smaller, which meant the soil released less water per unit change in matric

suction. As shown in Figures 4.4, the treatments that caused swelling and dispersion all flattened

the water retention curves, so they all released less water per unit change in suction than the soils

treated to keep their water retention curves steeper.

i)

97

Table 4.4. Integral water capacity, IWC, of each soil horizon after leaching with solutions of differing salinity and sodicity. IWC calculated using different weighting functions. Initial EC for each horizon indicated in parentheses (dS m-1). Shadings in 3rd & 4th columns indicate treatments where data were combined to form single, average water retention or soil resistance curves.

Soil depth cm

Treatment No. of WRCs

combined

No. of SRCs

combined

PAW, mm/m

IWC, mm/m, with individual & combined weighting functions Osmotic

stress only

Soil resistance

only

Aeration only

Hydraulic conductivity

only

All weightings

applied

0 –

10

Initial EC (7.56)

6 6 290

122

272 221 262

86 0.1 M CaCl2 256 129 RO water 250 127 1/2 initial EC 156 97 1/4 initial EC 196 110 1/16 initial EC 247 126

10 -

25

Initial EC (12.29) 2

6

153 46 129 103 148 30 0.1 M CaCl2 134 74 RO water

4 180

160

146 77 168

50 1/2 initial EC 96 38 1/4 initial EC 113 42 1/16 initial EC 146 48

25 -

35

Initial EC (13.90) 2 2 166 37 136 131 160 25 0.1 M CaCl2 139 86 RO water 1

4

142 132 117 48 134 31 1/2 initial EC

3 154 75

126 97 141 39

1/4 initial EC 98 48 1/16 initial EC 131 61

35 -

55


4

159 154 138 65 137 36 1/2 initial EC

3 192 120

168 101 168 58


55 -

75


4

211 200 104 81 186 41 1/2 initial EC

3 223 142

125 104 201 46


75 -

100


4

189 168 146 77 164 38 1/2 initial EC

3 227 149

178 134 197 62


100

- 115

Initial EC (8.89) 2 2 162 74 124 95 149 34 0.1 M CaCl2 145 57 RO water 1 1 137 121 106 25 116 8.6 1/2 initial EC 2 3 150 105 116 53 135 25 1/4 initial EC 121 26 1/16 initial EC 1 157 149 114 60 135 27

115

- 150

Initial EC (9.03) 2 2 164 74 107 114 151 29 0.1 M CaCl2 147 58 RO water 1 1 158 138 132 33 - - 1/2 initial EC 2 3 185 126 148 65 171 30 1/4 initial EC 148 32 1/16 initial EC 1 188 178 138 65 165 27

> 15

0


3

109 96 91 5.4 99 1.1 1/2 initial EC 2 153 115 126 35 137 0.017 1/4 initial EC 129 0.018 1/16 initial EC 1 158 152 124 35 135 0.014

98

The application of various weighting functions to account for high salinity, poor soil aeration, high

soil penetration resistance, and low hydraulic conductivity altered the amount of available water

significantly. Examples are shown for the soil when it was leached with a salt solution having the

same chemistry as its initial field state in Figure 4.6. The profile of solid red triangles shown on the

far left-hand side of Figure 4.6 represents the most extreme weighting of the water capacity, while

the open circles with the dashed black line on the far right-hand side of Figure 4.6 represent the

classical PAW, which puts no weighting on the water capacity. The lines in between the two

extremes show profiles of available water where the water capacity was weighted for the individual

factors; the results are consistent with the findings of Chapter 2, namely, that the weighting for high

salinity is the most severe, followed by that for poor soil aeration, then high soil penetration

resistance, then least significant of all the weightings, the unsaturated hydraulic conductivity.

Figure 4.6 Profiles of soil water availability calculated by weighting the water capacity of the soil in its initial saline state for different limiting factors.

99

Table 4.4 shows that the magnitude of PAW was always greater than any IWC values; this, of

course, is because PAW is not attenuated in any way – it represents the maximum possible amount

of water that can be extracted from the soil. So any attenuation of the water capacity must

obviously reduce the amount of available water. As found in Chapter 2, weighting of the water

capacity for salt (osmotic stress) caused the greatest single reduction in available water of any of

the weighting factors. Groenevelt et al. (2004) acknowledged that this weighting function for

osmotic stress creates the most severe attenuation possible because it assumes plants completely

exclude all salt during water uptake, which of course is not true. They do not behave like perfect

osmometers (Bazihizina et al. 2012) – they actively adjust their root cell membranes to allow (or

prevent) cations and anions to be taken up or released as required, a process known as osmo-

regulation. The extent to which plant cell membranes ‘open’ or ‘close’ in response to osmotic

stresses is quantified using a reflection coefficient, �. Groenevelt et al. (2004) argued that, as a first

approximation, plants osmo-regulate by adjusting � such that the product of the osmotic head of

the soil solution and the reflection coefficient maintains a relatively constant cell turgor pressure, T,

viz.

T = ho x �(t, hm, ho) , [4.2]

where �(t, hm, ho) is a complex plant-specific function of time, soil matric and osmotic heads.

In the absence any real data on plant-reflection coefficients Groenevelt et al. (2004) chose a value

of � = 1, which implied that plants experience the full effect of any osmotic stress in the soil

solution. Given that different plants can vary their reflection coefficients in the range 0 < � < 1 over

time, it is likely that better estimates of IWC could be obtained by adjusting � downward in some

appropriate way. Exploring the nature of � for different plants under different soil and

environmental conditions is the subject of an entirely different research project, which is beyond

the scope of this thesis. The recent work of Bazihizina et al. (2012), however, provides useful data

that might be used in such an analysis.

Despite the severity of the osmotic attenuation applied above, the reclamation process of reducing

salinity and sodicity caused an increase in IWC even when salinity effects alone were considered

using Groenevelt et al.’s (2004) model – i.e. no other attenuation applied (Figure 4.7a). When all

weighting functions were applied (including salinity effects) the recovery of IWC during

reclamation was considerably less (Figure 4.7b). This again reflects the fact that the method to

account for salinity effects was too severe and needs to be adjusted using appropriate reflection

coefficients for different plants.

100

Figure 4.7 Increases in plant available water (IWC) during reclamation of the soil profile from its initially sodic-saline state to a calcic non-saline state calculated using a) only the osmotic weighting function of Groenevelt et al. (2004), and b) all weighting functions. NB. The scales on the available water axis for parts a) and b) are different.

a)

b)

101

4.4 Conclusions

Although it was already known from Jayawardane and Beattile’s (1979) work that the water

retention curve of a saline-sodic soil can be shifted by changing the composition of the soil solution

toward calcic and non-saline, the work of this chapter has demonstrated that the changes to water

retention are accompanied by changes in other soil physical properties that have consequences for

plant available water.

Under the conditions established in this experiment, it was possible to make modest changes in the

structure of a saline/sodic soil profile (with no mechanical disturbance) by leaching it with a

relatively concentrated solution of calcium. Although the soil penetration resistance tended to be

higher in the calcium-treated soils than other treatments (because they drained faster and were drier

at most matric heads), the water retention curves were steeper than the other treatments, which

promotes good aeration. Furthermore, once the calcium salt was leached out, the salinity dropped

and thus the plant available water (as measured by IWC) was higher than the other treatments.

By simply leaching out the high SAR soil solution from the soil using progressively more dilute

solutions, the soil experienced greater and greater swelling and dispersion, which shifted the pore

size distribution to finer and finer pores. This had the effect of making the water retention curves

flatter and thus more poorly aerated at most matric heads. Plant available water, as measured by the

IWC generally declined as swelling and dispersion progressed during leaching, even though the

penetration resistance declined and thus made the soil softer and easier to penetrate. This small

advantage was more than compensated for by the poor soil aeration that accompanied the excessive

swelling and dispersion in these soils.

While the slightly lower bulk density of the test soil cores may have influenced the magnitude of

some soil physical properties and their changes during reclamation, I believe the nature of the

changes is likely to be reflected in undisturbed field soil. The soil properties that had the greatest

limiting effect on soil water were (in order of declining importance: salinity >> aeration > soil

resistance > hydraulic conductivity. It is likely, however, that if more realistic reflection

coefficients can be devised to reduce the impact of salinity on IWC, the limiting role of poor soil

aeration will be shown to be considerably more important by comparison. It could be particularly

important in the lower horizons, where soil aeration is restricted by high bulk density.

102

Chapter 5 Shape of the salinity weighting function, ��o(h), based upon early plant response to

osmotic and matric stresses

5.1 Introduction

Many studies have shown that a wide variety of non halophytic plants struggle to survive in saline

soils (Bernstein and Hayward 1958; Hoffman and Rawlins 1971; Bernstein 1975; Maas and

Hoffman 1977; Grieve and Maas 1988). Most can tolerate a small amount of salt but there is a

threshold concentration above which growth declines in a linear fashion. The magnitude of this

threshold concentration (EC) and the slope of the decline in growth varies with plant species (Maas

and Hoffman (1977). Stepphun et al. (2005) suggested that the decline in plant growth occurs in a

more continuous, S-shaped fashion rather than abruptly like a bent-stick. Furthermore, Kopittke et

al. (2009) found that plant growth of non halophytes is retarded even at very low concentrations

and that a ‘threshold concentration’ of salt may not apply to some plants.

Groenevelt et al. (2004) suggested an attenuation method to calculate the effect of osmotic stress

on soil water availability. Their attenuating function presented the worst-case scenario for water

availability by assuming that plants behave just like perfect osmometers and exclude all salt from

the soil solution. They acknowledged that this assumption was not very realistic and that plant cell

membranes exhibit a ‘reflection coefficient’, � (varying between 1 and 0) to adjust the extent to

which salts are excluded during water uptake. In their work, Groenevelt et al. (2004) used a

reflection coefficient of � = 1 (total salt exclusion) to demonstrate the principles involved but also

because they lacked information on the real magnitude of � for different plants. They pointed out,

however, that the classical PAW model (and all other models for that matter) use a default value of

� = 0, which essentially implies that plants completely ignore all salt in the soil water. Neither of

these extreme �-values is particularly close to reality but there was no published literature on such

things at the time of publication, so they simply set � = 1 to show the lower limits of water

availability when salt is taken into account.

As the work reported in Chapter 3 indicated, application of the Groenevelt et al-weighting function

for salt produced estimates of plant available water that were only half the true quantity of water

used by Rhodes grass in the field. That is, the weighting function was too severe and represented

the absolute minimum in available water accessible to a very salt-sensitive plant. There remains no

concrete evidence in the literature to guide the preparation of appropriate weighting functions for

salt in soil water, so an attempt was made here to gather such evidence experimentally. The

purpose of this study was therefore to prepare a weighting function for the water capacity

103

completely independent of the theory of Groenevelt et al (2004) and based upon plant behaviour

under varying osmotic and matric stresses. In effect, I started with the research question: What is

the shape of the function describing plant response to increasing salt concentration in soil? In

particular, is the conventional bent-stick model too simple, and is there a common shape of

weighting function for a range of different plants? I hypothesized an S-shaped curve would

describe the effect of salt on plant growth and that a bent-stick model would be too crude but

acknowledged that plant variability might make a bent-stick or other approach more acceptable.


The growth of two plant species was evaluated in both soil and solution culture using a

combination of 40 different osmotic, matric, plus osmo-matric (total) hydraulic potentials (Table

5.1). Different quantities of sodium chloride were mixed with the soil (in a large cement mixer,

Supermixer 2.2, for 15 minutes) to produce 6 different combinations of osmotic and matric

potential at each (approximate) total potential. The EC was measured either directly (solution

culture) or on 1:5 soil:water extracts. The measured values of EC1:5 were converted to ECe using

the method of Slavich and Petterson (1993) and the osmotic potential (mbar or cm) was calculated

from the relation, hos = 360 ECe (Richards 1953). Values for the matric potential were fixed in

advance and the sum of the two recorded as the total (osmo-matric) potential (Table 5.1). For the

experiments conducted in soil, there were 30 different total hydraulic potentials x 2 plants x 3

replicates = 180 experimental pots of soil. For the solution experiments there were 10 different

total hydraulic potentials x 2 plant species x 3 replicates = 60 experimental pots. In total the

experiment included 240 experimental pots in a completely randomized design.

For the 180 x soil pots, approximately 980 g of non-saline soil of loamy sand texture from

Monarto, South Australia (Chittleborough et al. 1976) were passed through a 5 mm sieve, and

packed to a bulk density of 1.40 g cm-3 into 700 mL white plastic pots (120 mm deep and tapered

120 mm diameter at top/100 mm diameter at bottom) with drainage holes. Matric potentials were

established and maintained daily by weight based upon the water retention curve prepared for this

soil packed to the same bulk density. The water retention model of Groenevelt et al. (2004),

anchored at both saturation and wilting points, was used for this purpose (Figure 5.1). To assist

with uniform water/solution distribution in this coarse textured soil, three x 2 mm diameter,

perforated plastic tubes (150 mm long) were installed vertically into each pot at equidistant points

in a 100 mm triangular arrangement. When water/solution was needed, it was added into the 3

vertical tubes using a syringe. Five seeds of Faba bean (Vicia faba cv. Fiord) and five seeds of

Rhodes grass (Chloris gayana cv. Pioneer) were germinated and transplanted into the soil. After

104

plants were established, a thick layer of gravel (3-6 mm diameter) was poured onto the soil surface

to minimize evaporation.

Table 5.1 Numbered list of osmotic and matric potentials/heads used in soil and solution culture. Osmotic potentials were calculated from the ECe values.

Soil culture Solution culture

Pot N

o. ECe Osmotic Matric Total

Pot N

o. ECe Osmotic Matric Total

dS m-1 Potential (-mbar) Head (cm) dS m-1 Potential (-mbar)

Head (cm) 1 0.881 317 250 567 31 0.985 355 0 355 2 1.466 528 200 728 32 1.550 558 0 558 3 1.789 644 150 794 33 2.450 882 0 882 4 1.997 719 100 819 34 3.380 1,217 0 1,217 5 2.300 828 71 899 35 6.820 2,455 0 2,455 6 2.471 889 71 960 36 9.770 3,517 0 3,517 7 0.881 317 500 817 37 13.200 4,752 0 4,752 8 1.734 624 400 1,024 38 16.770 6,037 0 6,037 9 2.259 813 300 1,113 39 30.500 10,980 0 10,980 10 2.818 1,015 200 1,215 40 44.100 15,876 0 15,876 11 3.416 1,230 100 1,330 12 3.486 1,255 71 1,326 13 0.881 317 750 1,067 14 1.950 702 600 1,302 15 2.888 1,040 450 1,490 16 3.433 1,236 300 1,536 17 3.990 1,436 150 1,586 18 4.792 1,725 71 1,796 19 0.881 317 1,000 1,317 20 2.543 915 800 1,715 21 3.334 1,200 600 1,800 22 4.119 1,483 400 1,883 23 5.314 1,913 200 2,113 24 5.865 2,111 71 2,182 25 0.881 317 2,000 2,317 26 3.404 225 1,600 2,825 27 4.816 734 1200 2,934 28 6.890 2,480 800 3,280 29 9.339 3,362 400 3,762 30 11.325 4,077 71 4,148

105

For the 60 x solution pots, each pot was filled with 3-6mm gravel to form a support-matrix into

which the plants could grow. Groups of 6 pots (2 plants x 3 reps) were placed into 10 litre

rectangular tubs (430 mm x 32 mm x 115 mm) containing 7 L of salt solution corresponding to

each osmotic potential. Dilute, water soluble fertilizer (Yates “Thrive”) was added to the solution at

a rate of 1.8 g/L and taken into account in calculating the osmotic potential; the EC of the solution

was monitored throughout the experiment. Five germinated seeds of the same plants were placed

carefully into the gravel in the different salt solutions. The solutions were refreshed every second

day to prevent algal growth and build up of salt.

All seedlings in both soil or solution culture were grown for 15 days after germination in

Glasshouse No.7 (Plant Research Centre, Waite Campus), which was set up to maintain appropriate

light, temperature and ventilation. At 15 days, all plants (roots and shoots) were harvested and

dried for 5 days at 65 C to measure dry matter production.

Figure 5.1 Soil water retention curve of Monarto soil packed at a bulk density of 1.4 g cm-3. Parameter values for the Groenevelt et al. (2004) equation are: �s = 0.405, �wp = 0.100; k0 = 0.409; k1 = 0.328, and n = 0.646.

106


5.3.1 Dry matter yield as a function of osmotic-, matric- and total water potential

To prepare a weighting function for the differential water capacity in terms of the matric potential,

it was important first to establish the extent to which the water stress experienced by plants under

osmotic stress alone was equivalent to that experienced by plants under matric stress alone and

under both osmo-matric stresses at the same time. The experiments reported here were ideally set

up to make such a comparison because the range of osmotic solution potentials enveloped the range

of matric and osmo-matric potentials to which the plants were exposed.

Figure 5.2 shows the mean whole-plant dry matter yield of Faba bean (Vicia faba cv. Fiord) and of

Rhodes grass (Chloris gayana cv. Pioneer) grown in either soil media or solution media ranging in

total hydraulic potential from 0 to nearly -16 bar. As it turned out, neither Faba bean (Vicia faba cv.

Fiord) nor Rhodes grass (Chloris gayana cv. Pioneer) produced any dry matter whatsoever beyond

-11 bars, so there were no points between the total hydraulic potentials of -11 and -16 bar.

Identifying a definitive potential at which death occurred due to salt in these experiments was

therefore not possible, so the final potential was arbitrarily set at -15 bar to enable the theory

presented by Groenevelt et al. (2004) to be relied upon for the initial part of this analysis.

Figure 5.2 Mean whole-plant dry matter yield per plant plotted as a function of the total hydraulic potential (absolute value) for Faba beans and Rhodes grass grown in solution culture and soil. The vertical bars represent standard deviations of each mean point.

107

Figure 5.2 suggests that although the standard deviations for the soil-grown plants (red points) were

considerably greater than those for the solution-cultured plants (blue points), both sets of data

effectively occupied a single curve for each plant species within the range of common potentials.

This important result concurs with the work of Wadleigh and Ayers (1945) despite the fact that we

know the effects of matric stresses are often more complex than osmotic stresses (due to their

effects on hydraulic properties(Marshall et al. 1996). However, because hydraulic properties are

accounted for separately in weighting the water capacity for the IWC, it was possible to evaluate

the effects of osmotic stress independently here. Furthermore, the plant-effects were not related to

Na-induced calcium deficiencies (suggested by Kopittke and Menzies (2005)) or any other nutrient

deficiencies because tissue analysis was performed and indicated no such problems. There was

some necrosis of older leaf tips and leaf margin bronzing in the plants exposed to the highest salt

concentration but these were not nutritional symptoms (Weir and Creswell 1994). The reduction in

dry matter with total hydraulic potential in this short-term study can therefore be attributed largely

to water stress rather than any nutritional or toxicity problems (Munns and Termaat 1986).

Another important feature of the data shown in Figure 5.2 is that the patterns of growth reduction

are essentially the same for both Faba beans and Rhodes grass (logarithmic reductions) despite the

fact that Rhodes grass is nominally salt tolerant while Faba beans are nominally salt-sensitive. For

the purpose of developing a salt-weighting function for the water capacity, it will therefore be

accepted that matric and osmotic potentials influence uptake of water by plant roots in essentially

the same way. This has implications for the way in which the weighting function was developed

below.

5.3.2 A plant-based weighting function to attenuate the water capacity

To produce a salinity weighting function based upon a plant-growth response, the average plant dry

weights shown in Figure 5.2, G(hom), were divided by the maximum average dry weight of plants

grown in the soil or solution having the lowest salt concentration, G0(hom). This produced relative

growth values, G′(hom), which are plotted in Figure 5.3 as a function of the total hydraulic potential,

hom. A suitable function to describe the G′(hom) data is not immediately obvious from the shape of,

and variability in, the data so the following functional form was (tentatively) adopted as a first

approximation and fitted by eye rather than by a least squares analysis:

( � ( ��) [5.1]

108

where G′max is the largest value of G′(hom), namely 1.0, which occurs in the near-saturated soil,

where hom � 0; G′wp is the smallest value of G′(hom), namely 0 or very close to it, which occurs at

the wilting point, hwp, calculated using Equation (12) of Groenevelt et al. (2004); ( and � are

adjustable parameters, the values for which are listed in Table 5.2.

Figure 5.3 Relative growth, G′(hom), as a function of the total hydraulic potential of the water for a) Faba beans, and b) Rhodes grass. The red-dashed line represents Equation [5.1], the parameters for which are given in Table 5.1.

109

Table 5.2 Parameters and constants for Equation [5.1] to describe the relative growth, G′(hom), of Faba beans and Rhodes grass.

Plant ( � G′max G′wp hwp Faba beans 3 0.8 1 0 15 bar Rhodes grass 2.5 0.7

Figure 5.3 shows that the red dashed line described by Equation [5.1] passes through the maximum

point (Matric potential = 0.0 bar, Relative dry matter yield G′(hom) = 1.0) and the minimum point

(Matric potential = 15 bar, Relative dry matter yield G′ (hom) = 0.0), which makes it an ideal

component in a weighting function, salt(hm), as follows:

*

*

"

, [5.2]

where hi is the initial matric potential at which salt begins to reduce water availability for the plant

(in this study, hi + 0), and hf is the final matric potential at which plants die from salt stress

(calculated from Equation (12) of Groenevelt et al. (2004), which provides the matric head, hm, at

which the total hydraulic head reaches 15 bar (wilting point). The fitting parameter, *, is a plant-

specific slope factor (e.g. the actual relative yield of each plant species under the specified

conditions), and ", is a soil specific slope factor (e.g. the actual, measured ECe of the saturated

soil). Two examples showing the variation in shape of the weighting functions described by

Equation [5.2] for Faba beans and Rhodes grass are given in Figure 5.4, and the parameters for all

40 possible combinations of salinity and plant response studied here are listed in Table 5.3.

To use the weighting function, the differential water capacity must be found; this comes from the

equation for the water retention curve of Figure 5.1:

� � , [5.3]

where �wp is the water content at the wilting point at |-hm| = 15 bar, k0, k1 and n are fitting

parameters.

110

Figure 5.4 Four examples of weighting functions (Equation [5.2]) to account for salt in the soil water for a) Faba Beans, and b) Rhodes Grass. The Blue, Green, Brown and Red lines are for Pots 1, 11, 23 and 37 respectively (colour-coordinated data highlighted in Table 5.3).

111

Table 5.3. Parameters from Equation [5.2] used in preparing a weighting function to attenuate the water capacity for salinity, based upon soil and plant factors combined. In this study, the initial onset of osmotic stress was deduced to occur from hi = 0.0025 bar for all examples. The value of hf in this table is the matric potential at which wilting occurs under the salinity conditions corresponding to the ECe; values were calculated from Equation (12) of Groenevelt et al. (2004). Colour shaded data are shown in Figure 5.4 above.

Pot No.

Soil factors Plant factor "" = ECe (dS m-1) hf (-bar) **Faba **Rhodes

1 0.881 13.762 0.627 0.660 2 1.466 12.994 0.625 0.822 3 1.789 12.580 0.665 0.585 4 1.997 12.323 0.623 0.734 5 2.300 11.956 0.615 0.576 6 2.471 11.754 0.732 0.878 7 0.881 13.762 0.683 0.542 8 1.734 12.649 0.555 0.694 9 2.259 12.005 0.562 0.757 10 2.818 11.354 0.674 0.791 11 3.416 10.694 0.626 0.527 12 3.486 10.619 0.673 0.806 13 0.881 13.762 0.596 0.530 14 1.950 12.380 0.662 0.615 15 2.888 11.274 0.576 0.756 16 3.433 10.675 0.610 0.698 17 3.990 10.094 0.577 0.642 18 4.792 9.146 0.593 0.695 19 0.881 13.762 0.557 0.537 20 2.543 11.670 0.562 0.575 21 3.334 10.782 0.545 0.673 22 4.119 9.965 0.566 0.681 23 5.314 8.837 0.563 0.726 24 5.865 8.361 0.557 0.721 25 0.881 13.762 0.518 0.480 26 3.404 10.706 0.560 0.748 27 4.816 9.290 0.558 0.668 28 6.890 7.544 0.546 0.661 29 9.339 5.914 0.546 0.495 30 11.325 4.871 0.457 0.474 31 0.985 13.65 1.000 1.000 32 1.550 12.77 0.875 0.979 33 2.450 11.56 0.677 0.867 34 3.380 10.35 0.518 0.775 35 6.820 6.39 0.514 0.618 36 9.770 3.915 0.487 0.566 37 13.200 2.155 0.405 0.352 38 16.770 1.22 0.360 0.177 39 30.500 0.2 0.301 0.087 40 44.100 0 0.000 0.000

112

One obtains the differential water capacity by differentiating Equation [5.3] with respect to hm:

� , [5.4]

The ‘effective’ water capacity, after attenuating the differential water capacity for salt, is obtained

simply by taking the product of C(hm) and the weighting function, salt(hm), viz.

, [5.5]

and finally the integral water capacity, IWC(hm) is obtained by integrating the effective water

capacity from its lower limit, hi, to its upper limit, hf:

, [5.6]

The right-hand segment of the unweighted water capacity for the soil used in this work is shown in

Figure 5.4 with corresponding segments of the weighted or ‘effective’ water capacities, E(hm),

superimposed for the conditions in Pots No.1, 11, 23 and 37 for Faba beans and Rhodes grass.

Figure 5.5 Differential water capacity (solid black line) with 4 x examples of effective water capacities superimposed for Pot 1 (solid blue line), Pot 11 (dashed green line), Pot 23 (dash-dot brown line), and Pot 37 (solid red line) for a) Faba beans and b) Rhodes grass.

113

Clearly, the attenuation of the differential water capacity is relatively small until the salt

concentration becomes quite severe; this is in contrast to the very severe attenuation applied by the

equation of Groenevelt et al. (2004). The more modest attenuation embodied in the plant-based

weighting function of Equation [5.2] is consistent with the severely underestimated amount of plant

available water predicted using the Groenevelt et al. (2004) model in Chapter 2 for the field-based

study. A comparison of the integral water capacities calculated by the different approaches for the

examples shown in Figures 5.4 and 5.5 can be made from the data in Table 5.4.

Table 5.4 Estimates of plant available water in soil of varying salinity based on soil properties, or a combination of soil properties and plant response for Faba beans and Rhodes grass.

Unit

PAW by classical method

IWC by theory of Groenevelt et al. (2004)

IWC by plant-based weighting mm/m

mm / m mm / m Faba beans Rhodes grass Pot 1 252 211 244 243 Pot 11 252 134 213 204 Pot 23 252 108 192 187 Pot 37 252 63 122 97

5.4 Conclusions

The theoretical work of Groenevelt et al. (2004) laid the foundations for an attempt to take into

account osmotic stresses that plants experience in saline soils. Assuming conservation of mass,

they argued that the salt concentration increased as the water content decreased, which enabled

them to link the salt concentration to the water retention curve, which was an important advance.

Importantly they argued a case for setting the so-called plant-cell reflection coefficient, � = 1.0,

which meant that plant cell membranes completely excluded all salts (analogous to a perfect

osmometer). This meant, of course, that all their estimates of plant available water in the soil

represented the absolute lower boundary, which they freely acknowledged because they had no

knowledge of the appropriate reflection coefficients to use. The approach taken by Groenevelt et al.

(2004) was to express the classical differential WATER capacity, C(h), as a differential soil

SOLUTION capacity, Com(hm), which they expressed for each salt concentration in terms of an

osmotic potential, ho, and plotted it as a function of the matric potential, hm. The work presented in

this Chapter was intended to evaluate just how severe Groenevelt et al.’s (2004) attenuation

function really was by growing two different types of plants in saline soils and solutions and

evaluating their relative dry matter yields. The dry matter yields were then normalised and used to

develop a weighting function for the standard differential water capacity, C(hm).

114

The plant responses were highly variable but generally predictable but it was very difficult to fit a

response curve to the results using simple functions. A rather unsatisfactory model was fitted to the

data for the purposes of developing a function that could be readily used to attenuate the water

capacity. To the extent that the plant-response model was acceptable or satisfactory, its use in

defining a weighting function was highly successful. The model contained parameters that can be

linked to soil and plant properties that would be relatively easy to test. The estimates of plant

available water derived from the plant-based study sit on the high side of those produced by the

model of Groenevelt et al. (2004), and suggest that there is sufficient variation between two plant

species on different salt tolerance that a wider range of plants should be examined to determine

how broadly applicable the proposed weighting function is.

115

Chapter 6 General discussion and directions for future research

6.1 Introduction

The work reported in this thesis was motivated by a desire to improve our ability to estimate the

amount of plant available water in soils beyond the classical methods enveloped in the terms “Plant

Available Water” (PAW) and “Least Limiting Water Range” (LLWR). It took the view that soils

can be considered to be water ‘capacitors’ that are influenced primarily by the physical properties

of the soil. The soil properties of particular concern in this work were the soluble salt concentration

in the soil water, poor soil aeration in wet soils, rising penetration resistance and declining

hydraulic conductivity in drying soils. Their effects on soil water availability were embodied in the

model proposed by Groenevelt et al. (2001; 2004) called the integral water capacity (IWC). The

theoretical framework for the IWC model is quite strong but there is little published evidence to

date to evaluate its integrity using real plants growing in real soils. From a personal perspective, I

come from the agricultural regions of the Mekong Delta, Vietnam, which is under threat from

salinization. I was therefore especially interested to evaluate the veracity of the Groenevelt et al.

(2004) model (which is not very intuitive!) to calculate plant available water on soils being

reclaimed from the saline-sodic state. The work reported in this thesis therefore aimed to address

three main questions and hypotheses:

Question 1 (basis for Chapter 2 work)

When soil salinity, aeration, strength and hydraulic conductivity are all taken into account, how

much soil water is available to nominally ‘salt-sensitive’ plants when calculated using the IWC

model of Groenevelt et al. (2004)? It was (null-) hypothesized that the variation in IWC caused by

the above soil properties would have minimal impact beyond what would be expected with natural

variability. The expected (alternative) hypothesis was that the IWC would be significantly reduced

in some way based upon the (predictable) effects of the relevant soil properties.


To what extent do the (laboratory-based) estimates of soil water availability using IWC coincide

with the (field-based) measurements of soil water use by real plants? It was (null-) hypothesized

that the laboratory estimates of IWC would match the water use by real plants in the field. The

alternative hypothesis was that there would be significant differences between the field and

laboratory estimates, presumably because certain factors were not taken into account.

116


When saline-sodic soils are ‘reclaimed’ toward the non-saline, calcic state, to what extent does soil

water availability change (in terms of IWC) without significant soil disturbance in the process? I

(null-) hypothesized that there would be no change in soil physical properties and IWC during

leaching without physical disturbance of the soil throughout the reclamation process – as happens

in the field. The alternative hypothesis was that significant changes in IWC would occur during

leaching and that no disturbance would be required to achieve such changes.


To what extent does the response of plants to increasing salt concentration mimic the peculiar

shape of the weighting function proposed by Groenevelt et al. (2004)? I (null-) hypothesized that

theory and reality would coincide, and that the Groenevelt et al. (2004) model would be confirmed.

The alternative hypothesis was that plants would behave quite differently from predictions based

on theory, and that the salinity impact on water uptake and plant growth would be much less severe

but would need to be characterized using some other weighting function based more closely upon

plant considerations.

6.2 Major findings (and future research)

The work to address Question 1 was achieved by taking undisturbed soil samples from the profile

of a saline soil whose texture gradually became heavier with depth. Water retention, soil resistance,

soil aeration and soil salinity were all measured and used to prepare appropriate weighting

functions to attenuate the differential water capacity and obtain different estimates of plant

available water down the soil profile. All weighting functions attenuated the water capacity and

reduced the IWC to varying degrees, each of which produced smaller estimates of plant available

water than the classical PAW model. Weighting due to salinity caused by far the greatest individual

reduction in IWC, followed by soil resistance, soil aeration, then hydraulic conductivity. The

combination of all factors, of course, reduced IWC the most. The null hypothesis can thus be

rejected and the alternative hypothesis tentatively accepted because there were clear effects of the

weighting functions; the alternative hypothesis is accepted ‘tentatively’, however, because a

statistical evaluation of the effects shown in Chapter 2 was not possible. Furthermore, many of the

weighting functions were applied with little or no knowledge of the real magnitude of their

parameters based upon real plant behaviour. It is known that some plants can survive in hard soil

better than other (Taylor and Ratliff 1969; Aggarwal and Prihar 1975; Clark et al. 1999), and some

plants grow well under water-logged conditions while others cannot. To take this into account,

117

weighting functions were proposed for each limiting soil property having functional forms that

included plant-specific parameters, whose magnitude can be varied widely for different plants. The

plant-specific parameters attenuate the water capacity severely when a plant species is sensitive to a

restricting soil property (e.g. soil strength, soil aeration, low hydraulic conductivity).

Future research directions: The magnitude of the plant-specific parameters, of course, was

unknown in this work, but the theoretical framework is now set for somebody to determine the

magnitude of these parameters using a wide range of different plants in either glasshouse or field

studies. The plant-specific parameter for high soil strength, � in Equation [2.3], varies in the range

0 < � < �m, where �m is the maximum value of � and must be established experimentally. The

plant-specific parameter for poor soil aeration, A in Equation [2.4], varies in the range 0 < A < Am,

where Am is the maximum value of A and must be established experimentally. Similarly, the plant-

specific parameter for declining hydraulic conductivity, � in Equation [2.9], varies in the range 0 <

� < �m, where �m is the maximum value of � and must be established experimentally.

The work to address Question 2 was conducted in the field on the same soil used to address

Question 1. A water budget was constructed by saturating 3 x isolated blocks of soil to a depth of

1.5 m and growing a crop of relatively salt-tolerant perennial Rhodes grass (Chloris gayana cv.

Pioneer) to full canopy coverage before stopping all water inputs. The volumetric water content

was monitored regularly (using a specially calibrated neutron moisture meter) as the crop

transpired water over several months until it eventually died from water stress. The total change in

water content down the profile was determined by the difference in water contents at the time of

saturation and those at the time of permanent plant wilting. The predicted and measured amounts of

available water were compared with the classical PAW model, and it was concluded that the

magnitude of attenuation proposed by Groenevelt et al. (2004) was too severe. Some effort was

made to adjust the plant-specific slope parameters, �, A, and �, but (as with the work in the

previous chapter) without any real knowledge of the magnitude of these parameters for different

plants, it was considered futile to expend much time adjusting the parameters without new

information about real plants.

Future research directions: The field work conducted in this chapter used a relatively salt-tolerant

plant rather than a salt-sensitive plant. It was considered more important to establish complete

canopy cover in this experiment (and thus ensure a successful field study) than to use a salt-

sensitive plant and achieve no germination and growth. The time required to repeat the work using

a salt-sensitive plant was prohibitive and therefore beyond the scope of this thesis. However, it will

be important to conduct the work using, at the very least, a salt-sensitive plant before concluding

118

that the Groenevelt et al. (2004) model produces an unrealistically severe attenuation of the water

capacity. Ideally such field work would be conducted on soils of varying salinity using plants

having a range of different salt tolerances. This would provide approximate magnitudes for the

plant-specific slope coefficients �, A and �. If it turns out to be too difficult to establish salt-

sensitive plants on saline soils, an alternative would be to establish the plants on a non-saline soil

and then apply saline irrigation water.

One other direction for future research in the field for this type of work is to evaluate the real limits

of water extraction from the soil – it is normal to assume upper and lower integration limits of

‘field capacity’ (10 kPa) and ‘permanent wilting point’ (1500 kPa). However, there is considerable

evidence to suggest that plants extract water from very wet soils (with matric heads smaller than 10

kPa) and from very dry soils (with matric heads greater than 1500 kPa). A combination of

measuring plant growth and monitoring the average soil water potential would produce more

realistic integration limits and thus go some way toward getting more accurate estimates of IWC.

The work to address Question 3 was conducted in the laboratory using re-packed soil cores

leached first with a saline solution (isotonic with field conditions) then with various different salt

solutions to determine the extent to which changes in the pore size distribution would be

accompanied by measurable changes in salinity, soil strength, hydraulic conductivity and aeration –

and thus, plant available water (IWC). Fifty-four different average water retention curves were

prepared in this experiment, and the curves were differentiated to produce water capacities that

were weighted according to procedures outlined in Chapter 2. As in Chapter 3, it was found that the

salinity weighting function caused the greatest reduction in IWC and was probably too severe. It

was also found that the other factors reduced the water capacity somewhat, in the declining order of

importance: salt > aeration > strength > hydraulic conductivity. It was a surprise to find that with

zero disturbance of the soil samples after packing them, the structure of the soil was able to be

changed to a small extent without disturbing it mechanically, simply by changing the composition

of the leaching solution.

Future research directions: Although it was possible to alter the structure and water retention

properties simply by leaching the soil columns with different solutions, the soil samples were really

only nominally ‘undisturbed’ because the soil was re-packed into the columns before leaching

began. To get a better feel for how this soil behaves with no disturbance during reclamation, it

would be important to obtain soil samples collected from the field in undisturbed soil cores.

119

One aspect of this work that remains unanswered is the extent to which the water retention curves

that were produced by leaching the saline soil with very dilute salt solutions (even distilled water)

could be recovered to their initial conditions by subsequently leaching them with calcium solutions.

It was possible to do this if the initially saline soil was leaching with calcium chloride, such that the

soil went from a flocculated, structurally stable state to another flocculated, structurally stable state

in one immediate step. It is less certain, however, whether structure once damaged by swelling and

dispersion, can be recovered back to its originally stable state simply through leaching. If it could

be shown that such a process is possible, there would be significant implications for the

reclamation of saline lands using minimum tillage schemes.

The work to address Question 4 was conducted to evaluate whether the very severe (theoretically

based) weighting function of Groenevelt et al. (2004) could be verified by plant observation. Plants

of two different types (Faba beans, Vicia faba cv. Fiord, and Rhodes grass, Chloris gayana cv.

Pioneer) were grown in a glasshouse in either pots of salt-solutions or in soil having different salt

concentrations. The idea was to develop a weighting function for salinity based upon measured

plant growth responses to varying levels of salinity, and compare this with the peculiarly shaped

weighting function for salt proposed by Groenevelt et al. (2004). The growth reduction pattern due

to salt was similar for both plants, so the relative growth of each plant was plotted as a function of

the total water potential. It was found that the relative growth of the solution-grown plants

coincided with those for the soil-grown plants, which implied the plants responded in the same way

to both osmotic and matric stresses. Relative growth responses were then fitted to a (rather

inadequate) model (Equation [5.1]), which was then used in a weighting function of the sort

proposed for other limiting factors. The weighting function included both plant- and soil-specific

fitting parameters that produced a much more gentle attenuation of the water capacity than the IWC

model of Groenevelt et al. (2004). This suggests there is considerable room to adjust the ‘reflection

coefficient’ in their model.

Plant responses to salinity have been investigated by many researchers on several plants and all of

them follow the classical ‘bent stick’ model, implying no plant response until the salt concentration

reaches a tipping point after which plant growth declines linearly to zero (Bernstein and Hayward

1958; Hoffman and Rawlins 1971; Bernstein 1975; Maas and Hoffman 1977; Grieve and Maas

1988). However, a more modest, smooth decline in plant growth with increasing salt concentration

is suggested by the present work.

Future research directions: The weighting function proposed in this chapter needs to be tested on

other plants and soils to determine whether the factors included in the fitting parameters of the

120

model are reasonably robust. The model suggested by Equation [5.1] for plant response does not

actually fit the data very well and so similar functions were used for both plants. In fact, the growth

response of the two plants was quite different, but there was insufficient time in this work to

explore other functions. A more accurate model based upon a more flexible function would

distinguish better between the species in terms of their response to salt.

121

Appendices

Appendix 1 Method used to prepare saturated paste extracts for analysis based upon Janzen

(1993).

Figure A1.1. Set-up used to obtain saturated paste extracts on which EC and pH were measured.

The gravimetric water content of air dried soil

was determined on samples ranging in weight

between 30 and 50 g. Separate samples of air-dry

soil were placed into containers with lids and their

weights recorded. (Depending on the volume of

solution extract required for analysis, the weight

of air-dry soil ranged between 200 and 400 g;

approximately one-third of the water added to

prepare the saturated paste could be recovered in

the saturation extraction process). Weighed

amounts of reverse osmosis de-ionized (RO)

water were mixed with the soil to bring it to saturation. Tests for complete saturation were

conducted as follows: at saturation, the soil paste glistened and flowed slightly when the container

was tilted on an angle; the paste slid cleanly from the spatula; when a trench was cut in the paste it

readily consolidated when the container was jarred. Samples were allowed to stand undisturbed and

covered for at least 4 h, then the above criteria for saturation were checked again. In cases where

free water ponded on the surface of the paste, additional air-dry soil was weighed-in and remixed.

In cases where the soil stiffened too much or did not glisten, some additional (weighed) RO water

was added and the mixing process repeated. With knowledge of the weighed amounts of soil and

water added, the water content of the saturation paste was determine, �SP. The saturation paste was

stood overnight to reach equilibrium, then re-mixed and pH was measured directly in the paste

122

(because this was more accurate than measuring the pH of the saturation paste extract). The paste

was then transferred to a Buchner funnel fitted with highly retentive filter paper; a suction was

applied using a vacuum pump to draw liquid out of the pastes; if colloids were present in the

filtrate, the filtering process was repeated until it was clear. The paste extracts were stored at 4oC

until analysed for EC and soluble cations and anions.

123

Appendix 2. Water retention curves for all 54 pots (6 treatments x 9 soil horizons) – each point is

an average of 4 points; vertical bars on each point represent ± 1 standard deviation of the mean..

Figure A2.1. Water retention curves for all 6 treatments of the soil horizon 0 to 10 cm.

124


125


126


127


128


129

Figure A2.7 Water retention curves for all 6 treatments of the soil horizon 100 to 115 cm.

130

Figure A2.8 Water retention curves for all 6 treatments of the soil horizon 115 to 150 cm.

131

Figure A2.9 Water retention curves for all 6 treatments of the soil horizon > 150 cm.

132

Appendix 3. Soil resistance curves for all 54 pots (6 treatments x 9 soil horizons) – each point is an

average of 4 points; vertical bars on each point represent ± standard deviation of the mean.

Figure A3.1 Soil resistance curves for all 6 treatments of the soil horizon 0 - 10 cm.

133


134


135


136


137


138


139


140

Figure A3.9 Soil resistance curves for all 6 treatments of the soil horizon > 150 cm.

141

References

Aggarwal GC, Prihar SS (1975) A simple technique to determine axial root growth force. Plant and Soil 42, 485-489. Ahmed S, Swindale LD, El-Swaify SA (1969) Effects of adsorbed cations on physical properties of tropical red earths and tropical black earths : I. Plastic limit, percentage stable aggregates, and hydraulic conductivity. Journal of Soil Science 20, 255-268. Akhter J, Murray RS, Mahmood K, Malik KA, Ahmed S (2004) Improvement of degraded physical properties of a saline-sodic soil by reclamation with kallar grass (Leptochloa fusca). Plant and Soil 258, 207-216. Allison IE (1965) Organic carbon. In 'Methods of Soil Analysis, Part 2, Chemical and microbiological properties' (Ed. CA Black et al.) pp. 1367-1378 (American Society of Agronomy: Madison,WI) Aly SM, Letey J (1990) Physical properties of sodium-treated soil as affected by two polymers. Soil Science Society of America Journal 54, 501-504. Amemiya M, Robinson CW, Cowley EW (1956) Reclamation of a saline-alkali soil in the upper Colorado River basin. Proceedings of Soil Science Society of America 3, 423-426. Angers DA, Carter MR (1996) Aggregation and organic matter storage in cool, humid agricultural soils. In 'Structure and Organic Matter Storage in Agricultural Soils.' (Eds MR Carter and BA Stewart) pp. 193-211. (Lewis-CRC Press: Boca Raton, FL). Armstrong W (1980) Aeration in higher plants. In 'Advances in Botanical Research.Vol. 7.' Ed. HW Woolhouse) pp. 225-332. (Academic Press). Arora HS, Coleman NT (1979) The influence of electrolyte concentration on flocculation of clay suspensions. Soil Science 127, 134-139. Asgarzadeh A, Mosaddeghi MR, Mahboubi AA, Nosrati A, Dexter AR (2010) Soil water availability for plants as quantified by conventional available water, least limiting water range and integral water capacity. Plant and Soil 335, 229-244. Bakker AC (1972) The comparative effect of exchangeable Ca, Mg, and Na on the stability in water of aggregates from the subsoil of red-brown earth. PhD Thesis. University of Adelaide, Adelaide. Baldock JA, Nelson PN (2000) Soil organic matter. In 'Handbook of Soil Science' (Ed. ME Sumner) pp. B33-B36. (CRC Press: London). Baver LD (1956) 'Soil Physics.' (John Wiley & Sons: New York). Bazihizina N, Barrett-Lennard EG, Colmer TD (2012) Plant growth and physiology under heterogeneous salinity. Plant Soil 354,1–19. Bell RW (1999) Boron. In 'Soil Analysis: An Interpretation Manual.' (Eds KI Preverill, LA Sparrow and DJ Reuter). pp. 309-316 (CSIRO publishing: Collingwood, Victoria).

142

Bernstein L (1975) Effects of salinity and socidity on plant growth. Annual Review of Phytopathology 13, 295-312. Bernstein L, Hayward HE (1958) Physiology of salt tolerance. Annual Review of Plant Physiology 9, 25-46. Blackmore AV, Miller RD (1961) Tactoid size and osmotic swelling in calcium montmorillonite. Soil Science Society of America Journal 25, 169-173. Blake GR (1965) Particle density. In 'Methods of Soil Analysis, Part 1, Physical and mineralogical properties, including statistics of measurement and sampling' (Ed. CA Black et al.) pp. 371-373 (American Society of Agronomy: Madison,WI). Blokhuis W (1980) Problem soils: their reclamation & management - Vertisols. In 'Land reclamation & water management: Development, problems & challenges.' IILRI Publication No.27, 43-72. (International Institute for Land Reclamation & Improvement: Wageningen, Netherlands). Bresler E, McNeal BL, Carter DL (1982) 'Saline and Sodic soils Principles-Dynamics-Modeling' (Springer-Verlag Berlin Heidelberg: New York). Bronick CJ, Lal R (2005) Soil structure and management: a review. Geoderma 124, 3-22.

Chahal SS (2010) Evaluation of soil hydraulic limitations in determining plant-available-water in light textured soils. PhD Thesis. University of Adelaide, South Australia.

Chan Y, Banin A, Borochovitch A (1983) Effect of potassium on soil structure in relation to hydraulic conductivity. Geoderma 30, 135-147.

Chhabra R (1985) Crop responses to phosphorus and potassium fertilization of a sodic soil. Agronomy Journal 77, 699-702. Chhabra R, Abrol IP, Singh MV (1981) Dynamics of phosphorus during reclamation of sodic soil. Soil Science 132, 319-324. Chittleborough DJ, Maschmedt DJ, Wright MJ (1976) Soils of the Monarto town site. Specific Land Use Survey Report No.SS-16. Department of Agriculture South Australia. Adelaide. 58p. Clark LJ, Bengough AG, Whalley WR, Dexter AR, Barraclough PB (1999) Maximum axial root growth pressure in pea seedlings: effects of measurement techniques and cultivars. Plant and Soil 209, 101-109. Cockroft B, Barley KP, Greacen EL (1969) The penetration of clays by fine probes and root tips. Australian Journal of Soil Research 7, 333-348. Connolly RD (1998) Modelling effects of soil structure on the water balance of soil–crop systems: a review. Soil and Tillage Research 48, 1-19. Cranner GR, Bowman DC (1991) Short-term leaf elongation kinetics of maize in response to salinity are independent of the root. Plant Physiology 95, 965-967.

143

Cresswell HP, Smiles DE, Williams J (1992) Soil structure, soil hydraulic properties and the soil water balance. Australian Journal of Soil Research 30, 265-283. da Silva AP, Kay BD, Perfect E (1994) Characterisation of the least limiting water range of soils. Soil Science Society of America Journal 58, 1775-1781. Davidson JL, Quirk JP (1961) The influence of dissolved gypsum on the pasture establishment on irrigated sodic clays. Australian Journal of Agricultural Research 12, 100-114. Day PR (1965) Particle fractionation and particle-size analysis. In 'Methods of Soil Analysis, Part 1, Physical and mineralogical properties, including statistics of measurement and sampling' (Ed. CA Black et al.) pp. 545-567 (American Society of Agronomy: Madison,WI) Delane R, Greenway H, Munns R, Gibbs J (1982) Ion concentration and carbohydrate status of elongating leaf tissue of Hordeum vulgare growing at high external NaCl. I Relationship between solute concentration and growth. Journal of Experimental Botany 33, 557-573. Dexter AR (1988) Advances in characterization of soil structure. Soil and Tillage Research 11, 199-238. Dikinya O, Hinz C, Aylmore G (2008) Decrease in hydraulic conductivity and particle release associated with self-filtration in saturated soil columns. Geoderma 146,192-200. Doneen LD, Henderson DW (1952) Soil conditions affecting infiltration of water and root development of crop plants. Proceedings of 7th General Meeting of American Society Sugar Beet Technologists, 7, 214-223. El-Morsy EA, Malik M, Letey J (1991) Polymer effects on the hydraulic conductivity of saline and sodic soil conditions. Soil Science 151, 430-435. El-Swaify SA, Ahmed S, Swindale LD (1970) Effects of adsorbed cations on physical properties of tropical red and tropical black earths : II. Liquid limit, degree of dispersion, and moisture retention. Journal of Soil Science 21, 188-198. Emerson WW, Chi CL (1977) Exchangeable calcium, magnesium and sodium and the dispersion of illites in water. II. Dispersion of illites in water. Australian Journal of Soil Research 15, 255-262. Emerson WW, Smith BH (1970) Magnesium, organic matter and soil structure. Nature 228, 453-454. Feddes RA, Kowalik PJ, Zaradny H (1978) 'Simulation of Field Water Use and Crop Yield.' Wiley & Sons. New York. 189p. Fitzpatrick RW, Boucher SC, Naidu R, Fritsch E (1994) Environmental consequences of soil sodicity. Australian Journal of Soil Research 32, 1069-1093. Flocker WJ, Vomocil JA, Howard FD (1959) Some growth responses of tomatoes to soil compaction. Soil Science Society of America Journal 23, 188-191. Gardner WR (1960) Dynamic aspects of water availability to plants. Soil Science 89, 63-73. Gardner WR, Nieman RH (1964) Lower limit of water availability to plants. Science 143, 1460-1462.

144

Golchin A, Baldock JA, Oades JM (1998) A model linking organic matter decomposition, chemistry and aggregate dynamics. In 'Soil Processes and The Carbon Cycle' (Ed. R. Lal, JM Kimble, RF Follett, BA Stewart) pp. 245-266. (CRC Press: Boca Raton). Grable AR (1966) Soil aeration and plant growth. Advances in Agronomy 18, 57–106 Grable AR, Siemer EG (1968) Effects of bulk density, aggregate size, and soil water suction on oxygen diffusion, redox potentials, and elongation of corn roots. Soil Science Society America Journal 32, 180-186. Grant CD, Groenevelt PH, Robinson NI (2010) Application of the Groenevelt–Grant soil water retention model to predict the hydraulic conductivity. Australian Journal of Soil Research 48, 447-458. Grant CD, Groenevelt PH, Semetsa S, Cass A (2003) Soil structure and available water in a fragile soil. 'Proceedings 16th Triennial Conference of International Soil and Tillage Research Organisation'. pp.498-504. Brisbane, Queensland. 13-18 July 2003. Greacen EL, Farrell DA, Cockroft B (1968) Soil resistance to metal probes and plant roots. Transactions 9th Congress of International Society of Soil Science. Vol. 1'. pp. 769-779. Greene RSB, Posner AM, Quirk JP (1978a) Interactions of suspensions of Ca-illite with PVA and Ca(OH)2. In 'Modfication of Soil Structure.' (Eds WW Emerson, RD Bond and AR Dexter) pp.157-163. (John Wiley & Sons: Bribane) Greene RSB, Posner AM, Quirk JP (1978b) A study of the coagulation of montmorillonite and illite suspensions using the electron microscope. In 'Modification of Soil Structure.' (Eds WW Emerson, RD Bond and AR Dexter). pp.35-40. (John Wiley and Sons: New York) Greene RSB, Rengasamy P, Ford GW, Chartres CJ, Millar JJ (1988) The effect of sodium and calcium on physical properties and micromorphology of two red-brown earth soils. Journal of Soil Science 39, 639-648. Gregory PJ (2006) 'Plant Roots: Their Growth, Activity, and Interaction with Soils'. Blackwell Publishing. Oxford. Grierson IT (1978) Gypsum and Red-brown Earth. In 'Modification of Soil Structure.' (Eds WW Emerson, RD Bond and AR Dexter). pp. 315-324. (John Wiley & Sons). Grieve CM, Maas EV (1988) Differential effects of sodium/calcium ratio on sorghum genotypes. Crop Science 28, 659-665. Groenevelt PH, Grant CD, Murray RS (2004) On water availability in saline soils. Australian Journal of Soil Research 42, 833-840. Groenevelt PH, Grant CD, Semetsa S (2001) A new procedure to determine soil water availability. Australian Journal of Soil Research 39, 577-598. Gupta RK, Singh RR (1988) A comparative evaluation of the reclamation efficiency of surface-concentrated versus internally incorporated calcium chloride and gypsum in an Alkali soil. Soil Science 146, 277-283. Gupta SC, Sharma PP, DeFranchi SA (1989) Compaction effects on soil structure. Advances in Agronomy 42, 311-338.

145

Hillel D (1971) 'Soil and Water: Physical Principles and Processes.' (Academic Press: New York). Hoffman GJ, Rawlins SL (1971) Growth and water potential of root crops as influenced by salinity and relative humidity. Agronomy Journal 63, 877-880. Hoffman GT, Rawlins SL, Garber MJ, Cullen EM (1971) Water relations and growth of cotton as influenced by salinity and relative humidity. Agronomy Journal 63, 822-826. Horn R, Baumgartl T, Kayser R, Baasch S (1995) Effect of aggregate strength on strength and stress distribution in structured soils. In 'Soil structure: its development and function.' (Eds KH Hartge and BA Stewart) pp. 31-52. (CRC Press, Inc.: Boca Raton, Florida) Ilyas M, Qureshi RH, Qadir MA (1997) Chemical changes in a saline-sodic soil after gypsum application and cropping. Soil Technology 10, 247-260. Isbell R (2002) 'The Australian Soil Classification.' (CSIRO: Melbourne) Janzen HH (1993) Soluble salts. In 'Soil Sampling and Methods of Analysis.' (Ed. MR Carter) pp. 161-166. (Lewis Publishers: Florida) Jayawardane NS, Beattie JA (1979) Effect of salt solution composition on moisture release curves of soil. Australian Journal of Soil Research 17, 89-99. Jayawardane NS, Chan KY (1994) The management of soil physical properties limiting crop production in Australian sodic soils--a review. Australian Journal of Soil Research 32, 13-44. Jeff B (1997) Soil and Water Relationships. http://www.noble.org/ag/soils/soilwaterrelationships/ (The Samuel Roberts Noble Foundation). Jensen CR (1982) Effect of soil water osmotic potential on growth and water relationships in barley during soil water depletion. Irrigation Science 3, 111-121. Joffe JS, Zimmerman M (1945) Sodium, calcium, and magnesium ratios in the exchange complex. Proceedings of Soil Science Society of America 9, 51-55. Jury WA, Jarrell WM, Devitt D (1979) Reclamation of saline-sodic soils by leaching. Soil Science Society of America Journal 43, 1100-1106. Kaddah MT, Fakhry SI (1961) Tolerance of Egyptian rice to salt. 1. Salinity effects when applied continuously and intermittently at different stages of growth after transplanting. Soil Science 91, 113-120. Kamphorst A (1990) Amelioration of sodic clay soils by crack stabilization: an experimental laboratory simulation. Soil Science 149, 218-227. Kay BD (1990) Rates of change of soil structure under different cropping systems. Advances in Soil Science 12, 1-52. Kemper WD, Koch EJ (1966) Aggregate stability of soils from western United States and Canada. USDA Technical Bulletin No.1355. US Government Printing Office. Washington DC.

146

Kinraide TB, Pedler JF, Parker DR (2004) Relative effectiveness of calcium and magnesium in the alleviation or rhizotoxicity in wheat induced by copper, zinc, aluminum, sodium, and low pH. Plant and Soil 259, 201-208. Koenigs FFR (1961) 'The mechanical stability of clay soils as influenced by moisture conditions and some other factors.' CLL Publication No.67-7. (Verslagen van landbouwkundige onderzoekingen: Wageningen). Koenigs FFR, Brinkman R (1964) Influence of partial sodium and magnesium saturation on the structural stability of clay soils. In 'The 8th International Congress of Soil Science,' 1964, Bucharest, pp. 219-225. Kopittke PM, Kopittke RA, Menzies NW (2009) Measurement and interpretation of salinity tolerance in four perennial grasses. Journal of Plant Nutrition 32, 30-43. Kopittke PM, Menzies NW (2005) Effect of high pH on Na induced Ca deficiency. Plant and Soil 269, 119-129. Krishnamurthy R, Anbazhagan M, Bhagwat KA (1987) Effect of NaCl toxicity of chlorophyll breakdown in rice. Indian Journal of Agricultural Science 57, 567-570. Lal R, Shukla MK (2004) 'Principles of Soil Physics.' (Marcel Dekker, Inc.: New York) 716p. Letey J (1985) Relationship between soil physical properties and crop production. Advances in Soil Science 1, 277-294. Lipiec J, Hakansson I, Tarkiewicz S, Kossowski J (1991) Soil physical properties and growth of spring barley as related to the degree of compactness of two soils. Soil and Tillage Research 19, 307-317. Maas EV, Hoffman GJ (1977) Crop salt tolerance - current assessment. Journal of the Irrigation and Drainage Division 103, 115-134. Marchuk A, Rengasamy P (2012) Threshold electrolyte concentration and dispersive potential in relation to CROSS in dispersive soils. Soil Research 50, 473-481. Marschner H (1995) 'Mineral Nutrition of Higher Plants.' (Academic Press: London, England) Marshall TJ, Holmes JW, Rose CW (1996) 'Soil physics' 3 edn. (Cambridge University Press: Cambridge). Martin JP, Ervin JO, Shepherd RA (1965) Decomposition and binding action of polysaccharides from Azotobacter Indicus (Beijerinckia) and other bacteria in soil. Proceedings of Soil Science of America 29, 397-400. Martin PK, Koebner RMD (1995) Sodium and chloride ions contribute synergistically to salt toxicity in wheat. Biologia Plantarum 37, 265-271. McGarry D (1990) Soil compaction and cotton growth on a Vertisol. Australian Journal of Soil Research 28, 869-877. McGeorge WT, Fuller WH (1950) Relation between the Na:Ca ratio in the saturation extract and the percentage sodium in the exchangeable complex. Transactions 4th International Congress of Soil Science 1, 400-403.

147

McIntyre DS (1958) Permeability measurements of soil crusts formed by raindrop impact. Soil Science 85, 185-189. Meiri A, Poljakoff-Mayber A (1970) Effect of various salinity regimes on growth, leaf expansion and transpiration rate of bean plants. Soil Science 109, 26-34. Mikhail EH (1974) Relationship between certain physical and chemical properties and subsoil friability in Lemnos loam. Australian Journal of Experimental Agriculture and Animal Husbandry 14, 694-697. Minasny B, McBratney (2003) Integral energy as a measure of soil-water availability. Plant and Soil 249, 253-262. Morgan JM (1984) Osmoregulation and water stress in higher plants. Annual Review of Plant Physiology 35, 299-319. Munns R, Schachtman DP, Condon AG (1995) The significance of a two phase growth response to salinity in wheat and barley. Australian Journal of Plant Physiology 22, 561-569. Munns R, Termaat A (1986) Whole-plant response to salinity. Australian Journal of Plant Physiology 13, 143-160. Neuman PM (1993) Rapid and reversible modification of extension capacity of cell walls in elongating maize leaf tissues responding to root addition and removal of NaCl. Plant, Cell and Environment 16, 1107-1114. Northcote KH (1979) 'A Factual Key for the Recognition of Australian Soils.' 4th edn. (Rellim Tech: Adelaide, South Australia) Northcote KH, Skene JKM (1972) 'Australian soils with saline and sodic properties.' (CSIRO Australia). Oades JM (1993) The role of biology in the formation stabilization and degradation of soil structure. Geoderma 56, 377-400. Oades JM, Waters AC (1991) Aggregate hierarchy in soils. Australian Journal of Soil Research 29, 815-828. Orcutt DM, Nilsen ET (2000) 'The Physiology of Plants Under Stress: Soil and Biotic Factors.' (John Wiley and Sons: New York). Oster JD (1993) Sodic soil reclamation. In 'Towards the Rational Use of High Salinity Tolerant Plants. Vol. 1.' (Eds H Leith and AA Masoom) pp. 485-490. (Kluwer Academic: The Netherlands). Oster JD, Shainberg I, Wood JD (1980) Flocculation value and gel structure of sodium/calcium montmorillonite and illite suspensions. Soil Science Society of America Journal 44, 955-959. Pasricha NS, Randhawa NS (1971) Available molybdenum status of some recently reclaimed saline sodic soils and its effect on concentration of molybdenum, copper,-sulphur and nitrogen in berseem (Trifolium alexandrinum) grown on these soils. Proceedings of International Symposium on Soil Fertility Evaluation. pp.1017-1026).

148

Patrick WH, Wyatt R (1964) Soil nitrogen loss as a result of alternate submergence and drying. Proceedings of Soil Science of America 28, 647-653. Pearson GA (1959) Factors influencing salinity of submerged soils and growth of Caloro rice. Soil Science 87, 198-206. Pearson GA (1961) The salt tolerance of rice. International Rice Communication Newsletter 10, 1-4. Peerlkamp PK (1950) The influence on soil structure of the “natural organic maturity” by roots and stubble of crops. Transactions 4th International Congress of Soil Science 2, 50-54. Qadir M, Noble AD, Shubert S, Thomas RJ, Arslan A (2006) Sodicity-induced land degradation and its sustainable management: Problems and Prospects. Land Degradation & Development 17, 661-676. Qadir M, Qureshi RH, Ahmad N (1996) Reclamation of a saline-sodic soil by gypsum and Leptochloa fusca. Geoderma 74, 207-217. Quirk JP (1986) Soil permeability in relation to sodicity and salinity. Philosophical Transactions of the Royal Society London 316, 297-317. Quirk JP (1994) Interparticle forces: A basis for the interpretation of soil physical behaviour. Advances in Agronomy 53, 121-183. Quirk JP, Schofield RK (1955) The effect of electrolyte concentration on soil permeability. Journal of Soil Science 6, 163-178. Rayment GE and Higginson FR (1992) 'Australian Laboratory Handbook of Soil and Water Chemical Methods' (Inkata Press: Melbourne) Rengasamy P (1983) Clay dispersion in relation to changes in the electrolyte composition of dialysed red-brown earths. Journal of Soil Science 34, 723-732. Rengasamy P (2002) Transient salinity and subsoil contraints to dryland farming in Australian sodic soils : an overview. Australian Journal of Experimental Agriculture 42, 351-361. Rengasamy P (2010) Osmotic and ionic effects of various electrolytes on the growth of wheat. Australian Journal of Soil Research 48, 120-124. Rengasamy P, Greene RSB, Ford GW, Mehanni AH (1984) Identification of dispersive behaviour and the management of Red-Brown earths. Australian Journal of Soil Research 22, 413-431. Rengasamy P, Marchuk A (2011) Cation ratio of soil structure (CROSS). Soil Research 49, 280-285. Rengasamy P, Olsson KA (1991) Sodicity and soil structure. Australian Journal of Soil Research 29, 935-952. Rengasamy P, Sumner ME (1998) Processes involved in sodic behavior. In 'Sodic soils: Distribution, Properties, Management, and Environmental Consequences.' (Eds ME Sumner and R Naidu). pp33-50. (Oxford University Press: New York)

149

Richards LA (1953) 'Diagnosis and improvement of saline and alkali soils.' USDA Handbook No.60. (United States Salinity Laboratory Staff: Riverside, CA). 160p. Richards LA, Wadleigh CH (1952) Soil water and plant growth. In 'Soil physical conditions and plant growth.' Ed. BT Shaw pp. 74-251. (Academic Press: New York). Robinson JB (1971) Salinity and the whole plant. In 'Salinity and water use.' (Eds T Talsma and JR Philip). pp.193-206. (The Macmillan Press: London). Rowell DL, Martin MW, Nye PH (1967) The measurement and mechanism of ion diffusion in soils III. The effect of moisture content and soil solution concentration on the self-diffusion of ions in soils. Journal of Soil Science 18, 204-222. Rowell DL, Payne D, Ahmad N (1969) The effect of the concentration and movement of solutions on the swelling dispersion and movement of clay in saline and alkali soils. Journal of Soil Science 20, 176-188. Shainberg I, Bresler E, Klausner Y (1971) Studies on Na/Ca, montmorillonite systems I. The swelling pressure. Soil Science 111, 214-219. Shainberg I, Letey J (1984) Response of soils to sodic and saline conditions. Hilgardia 52, 1-57. Shainberg I, Oster JD (1978) 'Quality of Irrigation Water.'IIIC Publication No.2.Volcani Centre. Bet Dagan. 65p Shainberg I, Shalhevet J (1984) 'Soil Salinity Under Irrigation: Processes and Management.' (Springer-Verlag: New York ) 349p. Sheldon AR (2009) Subsoil constraints to root growth and water use efficiency in northern grain soils: osmotic and toxic effects of salinity. PhD Thesis. University of Queensland, Brisbane. Sheldon AR, Menzies NW (2005) The effect of copper toxicity on the growth and root morphology of Rhodes grass (Chloris gayana Knuth.) in resin buffered solution culture. Plant and Soil 278, 341-349. Simmonds L, Mullins C, Schofield JL (1995) SPACTeach v1.0 MERTaLTM Courseware. In. ' Ed. SB Heath). (University of Aberdeen: Aberdeen, UK). Singh A, Chhabra R, Abrol IP (1979) Effect of fluorine and phosphorus applied to sodic soil on their availability and on yield and chemical composition of wheat. Soil Science 128, 90-97. Slavich PG, Petterson GH (1993) Estimating the electrical conductivity of saturated paste extracts from 1:5 soil, water suspensions and texture. Australian Journal of Soil Research 31, 73-81. Snedecor GW, Cochran WG (1989) 'Statistical Methods.' (Ames: Iowa State University Press). 503p. Steppuhn H, Van Genuchten MT, Grieve CM (2005) Root-zone salinity. II. Indices for tolerance in agricultural crops. Crop Science 45, 221-232. Sumner ME (1993) Sodic soils: new perspectives. Australian Journal of Soil Research 31, 683-750.

150

Taylor HM, Ratliff LF (1969) Root growth pressures of cotton, peas, and peanuts. Agronomy Journal 61, 398-402. Tisdall JM, Oades JM (1982) Organic matter and water-stable aggregates in soils. Journal of Soil Science 33, 141-163. van Hoorn JW (1958) Results of a groundwater level experimental field with arable crops on clay soils. Netherlands Journal Agricultural Science 6, 1-10. van Olphen H (1977) 'An introduction to clay colloid chemistry: for clay technologists, geologists, and soil scientists.' Second edn. (Wiley-Interscience: New York) 318p. Veihmeyer FJ, Hendrickson AH (1927) Soil-moisture conditions in relation to plant growth. Plant Physiology 1, 71-82. Veihmeyer FJ, Hendrickson AH (1949) Methods of measuring field capacity and wilting percentage of soils. Soil Science 68, 75-94. Verma S, Sharma PK (2008) Long-term effects of organics, fertilizers and cropping systems on soil physical productivity evaluated using a single value index (NLWR). Soil and Tillage Research 98, 1-10. Wadleigh CH, Ayers AD (1945) Growth and biochemical composition of bean plants as conditioned by soil moisture tension and salt concentration. Plant Physiology 20, 106-132. Warkentin BP, Schofield RK (1962) Swelling pressure of Na-montmorillonite in NaCl solutions. Journal of Soil Science 13, 98-105. Weber HW, van Rooyen PC (1971) Polysaccharides in molasses meal as an ameliorant for saline-sodic soils compared to other reclamation agents. Geoderma 6, 233-253. Weir RG, Creswell GC (1994) 'Plant Nutrient Disorders. 4. Pastures and Field Crops.' (Inkata Press: Melbourne). Wells T, Fityus S (2011) Neutron soil moisture probe operation in saline environments. Soil Science 176, 642-651. Wesseling JG, Stoof CR, Ritsema CJ, Oostindie K, Dekker LW (2009) The effect of soil texture and organic amendment on the hydrological behaviour of coarse-textured soils. Soil Use and Management 25, 274–283. Yeo AR, Lee KS, Izard P, Boursier PJ, Flowers TJ (1991) Short and long-term effects of salinity on leaf growth in rice (Oryza sative L.). Journal of Experimental Botany 42, 881-889. Zahow MF, Amrhein C (1992) Reclamation of a saline sodic soil using synthetic polymers and gypsum. Soil Science Society of America Journal 56, 1257-1260. Zaitseva RI, Sudnitsyn II (2001) The effect of mineral salts from soil solution on the water status and growth of barley seedlings. Eurasian Soil Science 34, 1240-1249. Zhang XC, Norton LD (2002) Effect of exchangeable Mg on saturated hydraulic conductivity, disaggregation and clay dispersion of disturbed soils. Journal of Hydrology 260, 194-205.

151

Zimmermann U, Schneider H, Wegner LH, Wagner HJ, Szimtenings M, Haase A, Bentrup FW (2002) What are the driving forces for water lifting in the xylem conduit? Physiologia Plantarum 114, 327-335.

Plant Availability of Water in Soils Being Reclaimed from the ...

Documents

Transcript of Plant Availability of Water in Soils Being Reclaimed from the ...