Printer Friendly

Size of subsoil clods affects soil-water availability in sand-clay mixtures.

Introduction

Texture-contrast soils (Isbell 2002) dominate a significant proportion of the cropping lands in the Mediterranean parts of southern Australian and present a peculiar set of soil physical and chemical problems in agriculture (Gardner et al. 1992; Hamblin et al. 1988; Harper and Gilkes 2004; Harper et al. 2000; Rebbeck et al. 2007). Typically in these soils, the sandy A horizon with <5% clay experiences severe water repellence at the soil surface, plus low fertility, and the clay-rich subsoil experiences poor soil structure. The two horizons are separated by a sharp boundary, which causes significant bypass flow (Tennant et al. 1992; Ritscma and Dekker 2000; National Committee on Soil and Terrain 2009; Hardie et al. 2011).

A common approach to ameliorate these soils is to add clay-rich material to the topsoil (Gardner et al. 1992; Tennant et al. 1992; Ward 1993; Cann 2000; Harper et al. 2000; Eldridge 2007; Hall et al. 2010; Betti et al. 2015), either by spreading or by delving (Desbiolles et al. 1997; Cann 2000; May et al. 2006; Hall et al 2009, 2010; Bailey et al. 2010; Davenport et al. 2011). Clay spreading serves to increase soil surface wettability, whereas delving modifies the entire soil profile by bringing up clay-rich subsoil clods and aggregates into the topsoil sand.

The average clay content of the topsoil increases by delving, but many of the clay-rich clods and aggregates remain discrete entities rather than becoming mixed intimately with the sand. An implication is that the affected soil volumes may continue to behave as unmodified pure sand in which plant roots can extract water and nutrients from the clay-rich clods only by growing mainly on or close to their surfaces. If such root growth behaviour is widespread in delved soils, it presents obvious difficulties for predicting the physical and chemical fertility of these soils, because this would require accurate description of the size and spatial distributions of the clay-rich clods and aggregates. This is not a trivial task in soils where the effects of delving operations vary significantly from place to place because of differences between operators and equipment as well as inherent differences in the soil profile being delved. Nevertheless, clay delving has become widespread on texture-contrast soils, so the water retention and transport properties of the root-zone in these highly modified soils need to be understood to enable yield predictions based on plant-available water (PAW).

If it were possible to predict soil hydraulic properties from the 'average' physical properties of uniform mixtures of sand and clay, plenty of literature is available on which to base such predictions (e.g. Rijtema 1969; Rawls et al. 1982; Brown et al. 1985; Gill et al. 2004; Fernandez-Galvez and Barahona 2005; Saxton and Rawls 2006; Lipiec et al. 2007; Martinez et al. 2008; Asgarzadeh et al. 2010; Costa et al. 2013). However, soils modified by clay delving arc far from texturally uniform; in fact, the clumps brought up into the root-zone range in size from tiny aggregates to very large clods as shown schematically in Fig. 1. The larger clods remain distinct from the sand and create a zone with bimodal soil physical properties rather than those of a natural soil having the same 'average' texture. Seeking guidance from the literature on the properties of natural soils having similar average texture can therefore lead to significant errors in prediction of PAW. The properties of a bimodal mixture must take into account the properties of the discrete materials involved. In this regard, the extent to which predictions of PAW could be based on bimodal v. uniform mixtures may depend on the size of the clay-rich aggregates. For example, smaller aggregates would be expected to produce physical properties that more closely resemble those of uniform mixtures, particularly when the aggregates of clay arc of similar size to the sand particles. Any dispersion of the aggregates would further the extent of mixing with sand. Larger clods would be expected to produce bimodal properties based upon the properties of the discrete materials in proportion to the respective quantities present in the mix. Only minimal effects of dispersion from the surfaces of large clods would be expected.

Theory

The schematic mixture shown in Fig. 1a suggests that the average volume of water ([m.sup.3]) in the mixed soil at a given soil matric head, [theta][(h).sub.mix] (h, in metres), is approximately the sum of the volumetric water contents ([m.sup.3] [m.sup.-3]) of the separate components multiplied by the respective volumes they occupy ([m.sup.3]):

[theta][(h).sub.mix] [V.sub.T] [equivalent to] [theta][(h).sub.C] [V.sub.C] + [theta][(h).sub.S] [V.sub.S] (1)

where [theta][(h).sub.mix], [theta][(h).sub.C] and [theta][(h).sub.S] are, respectively, the volumetric water contents of the mixture, the clay-rich aggregates and the sand at a given h: and [V.sub.T], [V.sub.C], and [V.sub.S] are their bulk volumes. Assuming that the clay-rich aggregates remain distinct from the sand, Eqn 1 can be rearranged to calculate the weighted average volumetric water content of the total mixture:

[theta][(h).sub.mix] [congruent to] [theta][(h).sub.C] x [V.sub.C / [V.sub.T] + [theta][(h).sub.S] x [V.sub.S] / [V.sub.T] (2)

The bulk volume each component occupies in the mixture, [V.sub.C] and [V.sub.S], is difficult to measure directly but can be estimated from their respective masses, [M.sub.C] and [M.sub.S] (kg), and bulk densities, [[rho].sub.C] and [[rho].sub.S] (kg[m.sup.-3]), which are relatively easy to measure independently. Thus:

[V.sub.C] = [M.sub.C] / [[rho].sub.C] and [V.sub.S] = [M.sub.S] / [[rho].sub.s] (3)

Now, defining PAW ([m.sup.3] [m.sup.-3]) as:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4)

where FC is field capacity (h = 1 m) and WP is the permanent wilting point (h = 150 m), the PAW for the total mixture of sand and discrete clay-rich aggregates shown in Fig. la, PA[W.sub.mix], can be estimated as:

PA[W.sub.mix] [congruent to] PA[W.sub.C] x [[V.sub.C] / [V.sub.T]] + PA[W.sub.S] x [[V.sub.S] / [V.sub.T]] (5)

Similarly, the bulk density of the total mixture in Fig. 1a. [[rho].sub.mix], can be calculated from the bulk densities of the distinct components, [[rho].sub.C] and [[rho].sub.S+] and their proportionate volumes in the total mixture:

[[rho].sub.mix] [congruent to] [[rho].sub.C [[V.sub.C] / [V.sub.T]] + [[rho].sub.S] [[V.sub.S] + [V.sub.T]] (6)

The extent to which Eqns 5 and 6 hold true for mixtures of sand and clay-rich clods or aggregates created during delving can be used to evaluate whether the delving operations significantly alter the 'texture' or whether the clay-rich aggregates remain distinct entities. This is important to understand, because if the clay-rich aggregates remain largely distinct and isolated (apart from minor dispersion around the external surfaces), any benefit of delving for crop production depends upon the probability of plant roots intercepting the clods. On the other hand, if the texture is more uniformly altered, plant roots can take advantage of potentially improved soil hydraulic properties with greater probability. Thus, if the components in the total mixture remain discrete as depicted in Fig. la, then Eqns 5 and 6 will hold true. If, on the other hand, a more intimate mixture occurs with the 'average' texture depicted in Fig. 1 b, then Eqns 5 and 6 will not hold true and the water-holding properties of the mixed soil will reflect the new structural arrangement of the intimately mixed components. The objective of this study was to evaluate the ways in which the size and quantity of clay-rich subsoil aggregates mixed with sand influence the amount of PAW as measured by the integral water capacity (IWC) (Groenevelt et al. 2001), which modifies the differential water capacity to account for physical limitations.

Materials and methods

Preparation of mixtures of sand and clay-rich subsoil aggregates

The sand and clay-rich subsoil aggregates were collected from texture-contrast soils at two agricultural sites in South Australia, near Coonalpyn (35[degrees]41 28 S, 139[degrees]53 05 E) and near Bordertown (36[degrees]12 52 S, 140[degrees]42 08 E). Both sites have shallow sandy topsoils (0.2 0.3 m deep) with an A2e bleached horizon over a sandy clay loam subsoil (Coonalpyn) or sandy clay subsoil (Bordertown) and they were classified as Brown Sodosols (Isbell 2002) or Stagnic Solonetzs (WRB IWG 2007). The sites were chosen as representative of the typical texture-contrast soils in the South East of South Australia that are clay-delved for the amelioration of their inherently poor productivity.

Soil dry bulk density was determined on undisturbed soil cores (0.05 m diameter, 0.05 m height) taken from the sand-textured A horizon and from the upper 0.2 m of the clay-rich B horizon near the maximum depth of tine-penetration during delving. Particle-size distribution (by mechanical separation; Smith and Tiller 1977) and pH and electrical conductivity (EC) in 1:5 soil:water suspensions (Rayment 1992) were determined on bulk samples taken beside the soil cores in each horizon (Table 1).

To prepare mixtures that simulated the schematic soils shown in Fig. 1a, b, we chose large and small aggregates from the B horizon to mix with the sand from the A horizon at both sites. From the Coonalpyn site, four different nominal sizes of clay-rich subsoil aggregates were gathered from the B horizon: <2 mm (passed through a 2-mm sieve), 6 mm (collected between sieves having grids of 4.75 and 6.7 mm), 20 mm and 45 mm (separated manually by measuring three orthogonal diameters). The subsoil clods from the Bordertown site were generally smaller than those from Coonalpyn, so only three clay-rich subsoil aggregate sizes were possible: <2, 6 and 20 mm (illustrated in Fig. 2). The different aggregate sizes were mixed with sand in weight/weight proportions of 10%, 20%, 40% and 60%, with the exception that, for the 45-mm aggregates from Coonalpyn, the smallest possible proportion was 20% (i.e. one clay-rich clod per core) and the largest proportion possible was 40% (two clay-rich clods per core).

Air-dried samples of the different size fractions of the clay-rich aggregates were mixed with the sand by gently folding the materials together on paper, dividing the mix into quarters, remixing and repeating three times. The clay content of subsamples of the different mixtures was determined by dispersion and sedimentation (Table 1) and nominal textural classes were assigned. The Coonalpyn mixtures were placed in stainless-steel rings of 5 cm height and 11 cm diameter (to accommodate clods having diameters up to 45 mm), and the Bordertown mixtures were placed in smaller stainless-steel rings of 5 cm height and 5 cm diameter. To allow consolidation of the mixtures in the rings, they were subjected to three preliminary cycles of wetting and air-drying (Lipiec et al. 2007; Shiel et al. 1988). The bulk density of each mixture was calculated from its total mass (corrected for the measured water content) and the volume occupied in its ring, taking into account any irregular shapes of the upper surface caused by protruding clay-rich aggregates or clods in the mixtures by weighing a quantity of very fine sand of known bulk density required to cover the protrusions. For samples comprising 40% and 60% of clay-rich aggregates <2 mm, some shrinkage occurred away from the edges of the rings after the preliminary wetting and drying, so the sample volumes were determined by removing them from their rings after all measurements were taken, sealing them in a paraffin coating and measuring their volumes by displacement in water.

Integral water capacity

Volumetric water retention ([m.sup.3] water [m.sup.-3] total) was measured at saturation plus seven different matric heads (m) by using saturated ceramic plates connected to hanging columns of water or placed in water-extraction chambers connected to pressurised [N.sub.2] gas until their weights did not change. Using the software Mathcad 14 (Parametric Technology Corporation 2007), the water retention data were fitted to the Groenevelt-Grant equation (Grant et al. 2010) anchored at the nominal wilting point in the relation:

[theta](h) = [[theta].sub.a] + [k.sub.1] x {exp [- [([k.sub.0] / [h.sub.a]).sup.n]] - exp [- [([k.sub.0] / h).sup.n]]} (7)

where [[theta].sub.a] and [h.sub.a] are, respectively, the volumetric water content and the matric head at the chosen anchor point a (permanent wilting point, [theta](150 m) in this case), [k.sub.1] and n are dimensionless fitting parameters and [k.sub.0] is a fitting parameter having units of the matric head (m). The differential water capacity, C(h) = d[theta]/dh, was determined as the first derivative of Eqn 7 and used to calculate the conventional PAW according to Eqns 4 and 5 ([h.sub.FC] = 1 m, [h.sub.PWP] = 150 m) and the IWC of Groenevelt et al. (2001):

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (8)

where [[omega].sub.i](h) are weighting functions (i = 1 to n) that have values ranging between 1 (no limitation to water availability) and 0 (complete limitation) to account for multiple possible soil physical restrictions that limit water availability. The soil restrictions for which weighting functions were applied in this study included poor soil aeration, [[omega].sub.A](h), limitations to plant water uptake due to rapid drainage (excessively large hydraulic conductivity, [[omega].sub.k-WET](h)) and excessively small unsaturated hydraulic conductivity, [[omega].sub.k-DRY](h) and increasingly large soil resistance to penetration, [[omega].sub.SR(h). The nature of the functions used was similar to those presented in Groenevelt et al. (2001) and Nang (2012). For example, the weighting function used to account for poor soil aeration was:

0 0 < h < [h.sub.minA]

[[omega].sub.A](h) {log(h/[h.sub.minA]) / log ([h.sub.minA] / [h.sub.minA]} [h.sub.minA] < h < [h.sub.maxA] 1 h < [h.sub.maxA] (9)

where [h.sub.minA] was the matric head corresponding to a volumetric water content [[theta].sub.minA] = [[theta].sub.s] - 0.1, where [[theta].sub.s] is the saturated volumetric water content, and 0.1 represents the minimum volumetric air content of 0.1 [m.sup.3] air [m.sup.-3] total required by many plants (da Silva et al. 1994); and [h.sub.maxA] was the matric head corresponding to a well-aerated soil with a volumetric water content [[theta].sub.maxA] = [[theta].sub.s] - 0.15, where 0.15 represents an adequate volumetric air content of 0.15 [m.sup.3] air [m.sup.-3] total (Groenevelt et al. 2001).

The weighting function used to account for increasing soil resistance was (Groenevelt et al. 2001):

1 0 < h < [h.sub.minSR]

[[omega].sub.SR](h) = 2.5 - SR(h) / 2 [h.sub.minSR] < h < [h.sub.maxSR] 0 h > [h.sub.maxSR] (10)

where [h.sub.minSR] represents the matric head above which the soil resistance begins to restrict root exploration of the soil and thus access to water (corresponding to SR(h) = 0.5 MPa), and [h.sub.maxSR] represents the matric head at which roots are completely prevented from exploring the soil and thus from taking up water (corresponding to SR(/t)=2.5 MPa), in accordance with evidence in the literature (Cockroft et al. 1969; Cockroft and Olsson 2000). To obtain the SR(h) function, soil resistance to penetration (SR, MPa) was measured on each sample in its ring at different matric heads over a period of months by using a LF-plus penetrometer (Lloyd Instruments, Bognor Regis, UK) with a 2.5-mm stainless-steel cone (30[degrees] angle) and a 2-mm-diameter recessed shaft advanced at 3 mm [min.sup.-1]. The average measured force (N) within the central 20-mm section of each core was converted to pressure using the cross-sectional area of the cone base, plotted as a function of the matric head, h, and fitted to a power function:

SR (h) = [sigma][h.sup.b] (11)

where the coefficients [sigma] and b are fitting parameters calculated using a Levenberg-Marquardt least-squared optimisation procedure in Mathcad 14.

For the weighting function accounting for the limitations to plant water uptake due to high and low soil hydraulic conductivity, we first estimated the unsaturated hydraulic conductivity function, K(h), for each sample. We used Rijtema's (1969) two functions (adapted from Gardner 1958) to create K(h) functions for soil in the 'wet' domain:

[K.sub.wet](h) = [K.sub.s] exp (-[alpha]h) (12)

and for the 'dry' domain:

[K.sub.dry](h) = a[h.sup.-1.4] (13)

where the values used for the fitting parameters, [alpha] ([m.sup.-1]) and a ([m.sup.2.4] d[ay.sup.-1]), depend on soil texture (published in Rijtema 1969). The saturated hydraulic conductivity, [K.sub.s], was measured on each sample, including the sand alone and the subsoil clay-rich aggregates alone (three replicates) by using a constant head method (Reynolds et al. 2000). Values for the coefficients [alpha] and a (Table 2) were selected empirically by matching the data collected on our mixtures (e.g. water retention) with those from soils of similar texture presented in Rijtema (1969), with the 'wet' domain set as the range of matric heads h = 0-1 m and the 'dry' domain was set as the range of matric heads h = 1 - 150 m. The complete K(/r) curves were consequently obtained using Eqns 12 and 13, thus defined:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (14)

The weighting function [[omega].sub.kWET](h), accounting for the restriction to plant water uptake due to rapid drainage of water in the wet range (Groenevelt et al. 2001), was:

[[omega].sub.kWET](h) = ([K.sub.r] (1) / [K.sub.r])(h)) 0 < h < 1 (15)

in which [K.sub.r](h) is the relative hydraulic conductivity, K(h)/[K.sub.s]. With no other information available, the range of matric heads at the wet end was arbitrarily set between h = 0 and 1 m. Similarly, [[omega].sub.kDRY](h), accounting for the restriction to plant water uptake due to low (declining) hydraulic conductivity in the dry end, was:

[[omega].sub.kDRY](h) = 1 - [[K.sub.r](h) / [K.sub.r](150) / 1 - [K.sub.r] ([h.sub.limK]) / [K.sub.r] (150) [h.sub.limK] < h < 150 m (16)

where [h.sub.limK] is the matric head from which the declining K(h) starts restricting water uptake to plants. From this point, we consider the unsaturated hydraulic conductivity to be so small that it is unable to deliver enough water to accommodate plant demand. The value of [h.sub.limK] depends on environmental conditions and plant species, and little published information is available to determine this point. We therefore applied the method proposed by van Lier et al. (2006), which determines the value of [h.sub.limK] (which those authors call 'limiting soil hydraulic condition') through a model based on transpiration demand (mm [day.sup.-1]) plus plant root density (m [m.sup.-3]) and the matric flux potential (M(h) [cm.sup.2] [day.sup.-1]) as a parameter to define hydraulic demand. (Note, however, that matric flux potential, M(h), is more correctly named 'matric flux transform' (see Grant and Groenevelt 2015).) van Lier et al. (2006) produced a nomogram to estimate the matric flux transform at which water availability starts to become limiting. For the purposes of this paper, a value of [M.sub.lim] = 0.1 [cm.sup.2] [day.sup.-1] was chosen on the nomogram in Fig. 3 to define the environmental conditions. For each soil [h.sup.lim.K] was calculated using the relation:

[M.sub.lim] = M([h.sub.limK]) = [[integral].sup.150.sub.h] K[([h.sub.limK])dh (17)

The integral water capacity, IWC, was calculated according to Eqn 8 using Eqns 9, 10, 15 and 16 as required to identify individual and overall effects of the physical restrictions on soil water availability. For mixtures containing large aggregates, their proportional effects on IWC were calculated according to the approach embodied in Eqns 5 and 6, namely:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (18)

As with Eqn 5, Eqn 18 relates to a bimodal soil mixture such as that shown in Fig. 1a, where [C.sub.C] (h) and [C.sub.S](h) are the respective water capacities of the clay-rich subsoil and the sand, and assumes (as also observed in the field) that plant roots grow primarily in the sand near the external surfaces of the subsoil aggregates rather than within them. On this basis, no physical limitations to plant water uptake from the subsoil aggregates need to be considered for [C.sub.C](h). By contrast, soil resistance and hydraulic conductivity were considered limiting factors in the sand, so their weighting functions, [[omega].sub.SR](h), [[omega].sub.KWET](h) and [[omega].sub.KDRY](h) all need to be considered for [C.sub.S](h) in Eqn 18. Aeration of the total mixture, however, needs to be accounted for in Eqn 18 by placing the weighting function, [[omega].sub.Amix](h) outside the brackets; in this context, [[omega].sub.Amix](h) uses the water-retention curve of the sand adjusted downward to remove the fractional volume of clay aggregates, [V.sub.C] [V.sub.T.sup-1], which does not contribute to aeration (so [[theta].sub.minA](h) of the mix equals [[theta].sub.minA] of the sand minus [V.sub.C] [V.sub.T.sup.-1]). Similarly, [[theta].sub.maxA] of the mix equals [[theta].sub.maxA] of the sand minus [V.sub.C] [V.sub.T.sup.-1].

Results and discussion

Bulk density and soil cohesion

The average bulk density of all mixtures for both the Coonalpyn and Bordertown sites increased with increasing amounts of clay-rich subsoil aggregates mixed with the pure A-horizon sand (Table 2). In accordance with the model proposed in Fig. la, the Coonalpyn samples containing large clay-rich aggregates (i.e. 20 and 45 mm) produced bulk densities that were equal to (or slightly greater than) that predicted by Eqn 6 with minor exceptions for the Bordertown samples. Similarly, the samples containing small clay-rich aggregates (i.e. <2 mm, as in Fig. 1 b) generally produced bulk densities that were less than, or equal to, that predicted by Eqn 6. Samples containing clay-rich aggregates of intermediate size (i.e. 6 mm) gave variable bulk densities, particularly with the lesser mix proportions (i.e. <20% by weight). However, for the greater mix proportions (i.e. 40% and 60% by weight), the bulk densities of the samples containing the intermediate 6-mm aggregates tended to behave like those having <2-mm aggregates, such that their bulk densities were less than, or equal to, that predicted by Eqn 6. This suggests that when sufficient quantities of clay-rich subsoil material are added to sand, the smaller the aggregates the more likely they are to form intimate mixtures having a lower average bulk density than their discrete components.

When lesser quantities of fine, clay-rich material are added to sand, the packing arrangements may not form intimate structural mixtures (as suggested by the bulk densities) but the cohesive behaviour is nevertheless affected. Figure 4 suggests that the addition of only 10% <2-mm clay-rich aggregates was sufficient to generate a firm, cohesive soil core, whereas the same quantity of intermediate (6-mm) or large (20-mm) clay-rich aggregates generated very little overall cohesion; the large aggregates simply fell away from the loose sand as soon as the confining ring was removed.

Soil resistance to penetration

For both soils, resistance to penetration as a function of soil matric head, SR(h), increased with increasing amounts of clay-rich subsoil aggregates in the mix, consistent with the power function shown in Eqn 11, the parameters for which are listed in Table 2. With the exception of the two treatments containing very large aggregates (20% and 40% of 45-mm clay-rich subsoil aggregates, Coonalpyn), soil resistance increased with decreasing size of clay-rich aggregates in the mixtures, especially with aggregates <2 mm at drier soil matric heads. The SR(h) functions showed that most of the mixtures, especially the Bordertown samples, had low penetration resistance across all but the very driest soil matric heads (h = 150 m); even at h=150m, however, penetration resistance rarely exceeded 2.5 MPa. Despite large variability for both Coonalpyn and Bordertown samples, the mixtures that contained larger aggregates tended to offer low soil resistance to penetration, primarily because the probability of the penetrometer encountering a clay-rich aggregate was low in these mixtures (i.e. small number of large aggregates in the mix). The lower probability of hitting an aggregate was reflected in the greater standard deviation of the mean soil resistance for the mixtures with large aggregates at most suctions (data not shown).

The initial and final matric heads, [h.sub.minSR] and [h.sub.maxSR], respectively, at which the value of SR(h) equalled 0.5 and 2.5 MPa for each soil mixture are reported in Table 3; these determined the transition points in the weighting function for high soil resistance (Eqn 10) used to calculate the integral water capacity, IWC.

Saturated hydraulic conductivity

As might be expected, the addition of increasing amounts of clay-rich aggregates to the sand reduced the mean saturated hydraulic conductivity, Ks, by several orders of magnitude for both the Coonalpyn and Bordertown soils (Table 2). Even as little as 10% clay-rich aggregates caused a significant reduction in [K.sub.s]. Consistent with the model depicted in Fig. 1, smaller clay-rich aggregates reduced the mean [K.sub.s] to a greater extent than larger clay-rich aggregates (Table 2). For the same quantity of clay added, the finer, more intimately mixed aggregates generated smaller pores in the mixtures, whereas the larger aggregates left significant regions of unadulterated sand with large pores to conduct water. The standard deviations of the mean [K.sub.s] values shown in Table 2 were lower for the samples containing <2-mm aggregates, confirming their greater uniformity of mixing relative to the larger aggregate fractions. The effect of the unsaturated hydraulic conductivity on soil water availability using IWC is outlined below in conjunction with the other limiting soil factors.

Water-retention curves and soil water availability

The water-retention curves for the mixtures of sand and clay-rich aggregates are shown in Fig. 5 (Coonalpyn) and Fig. 6 (Bordertown), and the fitting parameters in Eqn 8 for each curve are listed in Table 4. With the exception at the wet end in the Coonalpyn samples (0-0.5 m matric head), the curves for undisturbed pure sand and for undisturbed clay-rich subsoil formed an envelope surrounding the curves for all of the mixtures. As one might expect, the water-retention curves, without exception, moved away from the pure sand and upward towards the curves for pure clay-rich subsoil as the proportion of clay-rich aggregates in each mixture increased. In general, across the range of suctions between h = 0.1 and 1 m, the slope of the water-retention curves increased with increasing aggregate size in the order 2 mm <6 mm <20 mm <45 mm, with the only exception of the Bordertown mixtures at 60% subsoil content where the <2-mm treatments had the greatest slope (due to the much lower bulk density). The increase in slope with increasing aggregate size across this range of suctions shows that the smaller aggregates mixed intimately with the sand and created smaller pores to retain more water, whereas the larger aggregates remained discrete entities in the sand allowing for more large pores and lower water contents.

For the lowest proportions of clay-rich aggregates (10%) and the highest proportions (60%), only the smaller three aggregate sizes were available for comparison in the Coonalpyn soil (red, green and blue lines in Fig. 6a, d). The 10% clay-rich mixes (Fig. 5a) all behaved much like the pure sand regardless of aggregate size, and the 60% (Fig. 5d) clay-rich mixes behaved similar to one another and closer in shape to the undisturbed, clay-rich subsoil curve, regardless of aggregate size. For the Bordertown soil only three clay-rich aggregate sizes were used (<2 mm, 6 mm and 20 mm; larger aggregates were not available at this site), and their water retention curves followed essentially the same pattern as for the Coonalpyn soil mixtures, regardless of aggregate size.

Plant-available water and integral water capacity

Taking no account of physical limitations to water uptake from 'field capacity' (h = 1 m) to 'permanent wilting point' (h=150m), the PAW was low for the pure sands from Coonalpyn and Bordertown and significantly greater for the pure subsoil aggregates (observing the standard deviation bars in Fig. 7). This was in line with published values for similarly textured soils (Table 4). In general, there was no significant difference in PAW between the sands and the mixtures containing only 10% subsoil, regardless of aggregate size. However, PAW increased with increasing quantities of subsoil in the mixtures, and for any given proportion of subsoil, PAW tended to be greater for the mixtures containing smaller subsoil aggregates (Fig. 7). The lines for the different aggregate sizes are shown in Fig. 7 superimposed on the line produced by Eqn 5 for a bimodal mixture of discrete sand and subsoil. The close proximity of the lines for the 20-mm and 45-mm aggregates to that produced by Eqn 5 suggests that the larger subsoil aggregates existed as discrete entities in a 'matrix' of sand, whereas the finer (6-mm and <2-mm) subsoil aggregates combined with the sand to produce more intimate mixtures rather than bimodal mixtures. On this basis, subsequent evaluation of soil-water availability using IWC involved differing assumptions for the mixtures containing 'small' aggregates ([less than or equal to] 6mm) and 'large' aggregates ([greater than or equal to] 20mm), such that Eqn 18 was used to calculate IWC for the 'large' aggregates and Eqn 8 was used for the 'small' aggregates.

When poor soil aeration, high soil resistance, and high and low hydraulic conductivity are all included as limiting factors to soil-water availability (Table 5), the IWCs for the clay subsoils from both Coonalpyn (13 mm [m.sup.-1]) and Bordertown (24 mm [m.sup.-1]) were considerably lower than those for the pure sand at Coonalpyn (46 mm m ') and at Bordertown (64 mm [m.sup.-1]). The IWC of the subsoil clay was also lower than for any of the sand-clay mixtures at both sites. Separating the relative importance of the limiting factors can be achieved by examining the IWCs when individual limiting factors are considered on their own (Table 5). The IWCs were lowest for the subsoil clay when only the soil resistance was taken into account (IWC = 133 and 126 mm [m.sup.-1], respectively for Coonalpyn and Bordertown); high soil strength could therefore be argued to be the primary limiting factor in the pure subsoil clay. Poor soil aeration at the wet end and declining unsaturated hydraulic conductivity at the dry end were also important but they were comparatively less limiting. For the pure sands at both sites (IWC = 46 and 64 mm [m.sup.-1] at Coonalpyn and Bordertown, respectively), the primary limiting factor was the rapid drainage of water (large saturated hydraulic conductivity) in the wet range (Table 5). Hydraulic conductivity was also the main limiting factor even as the amount of subsoil clay in the mixtures increased, although the limiting effect diminished with decreasing size of the subsoil aggregates. Soil resistance was not a limiting factor at all in the pure sands and it only became somewhat limiting as large amounts of subsoil clay were added, especially in the smaller aggregate sizes.

Evaluating IWC (with all limiting factors taken into account) as a function of the amount of clay-rich subsoil in the mixture, Fig. 8a, b shows that soil-water availability generally increased with increasing additions of subsoil up to ~40%, after which the IWC plateaued. In fact, the solid and dashed black line fitted through the points predicted from Eqn 18 for bimodal mixtures (grey diamonds) suggests that greater additions of subsoil beyond 40-60% push the IWC downward towards that for the 100% clay-rich subsoil.

Additions of these large magnitudes, however, are unlikely to be economically viable in practice, so the results in the range 0-20% are of greater practical interest than those from the 40% and 60% additions. Additions of up to 20% Coonalpyn clay-rich subsoil, for example, suggest only minor if any advantage can be had from cultivating delved soil excessively to mix the subsoil with the sand (i.e. to produce aggregates [less than or equal to] 6 mm). This suggestion is borne out even more emphatically in the Bordertown soil where additions of 10% and 20% subsoil aggregates <6 mm appeared to depress the effects on 1WC relative to those produced by larger aggregates (predicted by Eqn 18). The practice of repeated cultivation after the initial delving to make it more uniform must therefore be questioned and should be evaluated in the field.

Conclusions

Mixtures of sand and subsoil clay were prepared in the laboratory from which water retention, soil resistance, and saturated hydraulic conductivity were measured. Similar studies have previously been conducted (Brown et al. 1985; Gill et al. 2004; Harper and Gilkes 2004), but to our knowledge, this was the first time that the sizes of the subsoil aggregates were taken into account when measuring the soil properties of the mixtures.

For both Coonalpyn and Bordertown soils, adding clay-rich subsoil to sand increased bulk density and soil penetration resistance, and reduced the saturated hydraulic conductivity of the mixtures towards those of the pure clay-rich subsoil. For extremely large additions of subsoil (i.e. 40-60%), the size of aggregates moderated the effects considerably. For example, small aggregates reduced the bulk density of the mixtures, presumably because they formed a more intimate mixture such that any swelling or dispersion of the clay created a greater total porosity between the sand particles. The greater porosity caused by the smaller aggregates, however, occurred at the expense of larger pores, so the hydraulic conductivity declined by at least an order of magnitude relative to that for the mixtures containing larger aggregates. Furthermore, smaller aggregates (even with only 10% subsoil) generated greater inter-particle cohesion (Fig. 4) and increased soil penetration resistance, similar to observations in analogous studies with homogeneous mixtures (Gill et al. 2004; Harper and Gilkes 2004).

The changes in physical properties generated by adding subsoil clay to sand also influenced the shape of the water-retention curves and the soil-water availability according to both PAW and 1WC models. This is an important outcome and it confirms the initial hypothesis that predictions of soil-water availability based merely on the average soil texture can lead to unrealistic estimates for clay delved or clay spread soils.

Addition of subsoil clay in the order of 10-20% significantly increased 1WC primarily by retaining more water and by reducing the excessively large hydraulic conductivity of the pure sand. In the practical range of clay additions (i.e. 10-20%), which corresponds to field quantities in the order of 150-330 t [ha.sup.-1] (depending on assumptions of depth of incorporation and average bulk density), there appears to be an effect of aggregate size. Larger aggregates and clods increased soil-water availability more so than smaller ones, implying little benefit in excessive post-delving cultivation to bring greater uniformity of the soil texture. The addition of 10-20% subsoil clay may seem quite a lot, but these additions are not unheard of in areas where clay delving is practiced. The apparent reversal in soil-water availability predicted for extremely large additions of subsoil clay (>40-60%) implies that excessive changes to soil profiles need to be considered carefully. The extent of crop yield variability on delved soils suggests that variations in post-delving tillage may be partly responsible for this, and should be evaluated quantitatively in the field.

Acknowledgements

The senior author (GB) was supported by a University of Adelaide Frederick James Sandoz Postgraduate Scholarship. The research was supported by a grant from the South East Natural Resource Management Board of South Australia.

References

Asgarzadeh H, Mosaddeghi M, Mahboubi A, 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. doi: 10.1007/s11104-010-0410-6

Bailey G, Hughes B, Tonkin R, Dowie R, Watkins N (2010) Gross soil modification of duplex soils through delving and spading. In '19th World Congress of Soil Science'. Brisbane. (International Union of Soil Sciences) Available at: http://iuss.org/19th%20WCSS/Symposium/pdf/ 0773.pdf

Betti G, Grant C, Churchman G, Murray R (2015) Increased profile wettability in texture-contrast soils from clay delving: case studies in South Australia. Soil Research 53, 125 136. doi: 10.1071/SR14133

Brown KW, Evans GB, Thomas JC (1985) Increased soil water retention by mixing horizons of shallow sandy soils. Soil Science 139, 118-121. doi: 10.1097/00010694-198502000-00004

Cann MA (2000) Clay spreading on water repellent sands in the south east of South Australia promoting sustainable agriculture. Journal of Hydrology 231--232, 333 341. doi:10.1016/S0022-1694(00)00205-5

Cockroft B, Olsson KA (2000) Degradation of soil structure due to coalescence of aggregates in no-till, no-traffic beds in irrigated crops. Australian Journal of Soil Research 38, 61-70. doi:10.1071/SR99079

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. doi: 10.1071/SR9690333

Costa A, Albuquerque JA, da Costa A, Pertile P, da Silva FR (2013) Water retention and availability in soils of the State of Santa Catarina-Brazil: effect of textural classes, soil classes and lithology. Revista Brasileira de Ciencia do Solo 37, 1535-1548. doi: 10.1590/S0100-0683201300060 0010

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. doi: 10.2136/sssaj1994.03615995005800060028x

Davenport D, Hughes B, Davies S, Hall D, and Rural Solution SA Agricultural Bureau of South Australia, Caring for our Country, Grains Research Development Corporation (2011) 'Spread, delve, spade, invert: A best practice guide to the addition of clay to sandy soils.' (Grains Research and Development Corporation: Kingston, ACT)

Desbiolles JMA, Fielke JM, Chaplin P (1997) An application of tine configuration to obtain subsoil delving for the management of non-wetting sands. In '3rd International Conference on Soil Dynamics (ICSD III)'. Tiberias, Israel, pp. 201-210. (Faculty of Agricultural Engineering, Technion--Israel Institute of Technology: Haifa, Israel)

Eldridge R (2007) Clay delving at Parilla. University of Adelaide Crop Science Newsletter Archive. Available at: www.adelaide.edu.au/css/ newsletters/archive/2000s/Eldridge_delving_2007.pdf

Fernandez-Galvez J, Barahona E (2005) Changes in soil water retention due to soil kneading. Agricultural Water Management 76, 53-61. doi: 10.1016/j.agwat.2005.01.004

Gardner WR (1958) Some steady-state solutions of the unsaturated moisture flow equation with application to evaporation from a water table. Soil Science 85, 228-232. doi:10.1097/00010694-195804000-00006

Gardner WK, Fawcett RG, Steed GR, Pratley JE, Whitfield DM, Van RH (1992) Crop production on duplex soils in south-eastern Australia. Australian Journal of Experimental Agriculture 32, 915-927. doi: 10.1071 /EA9920915

Gill JS, Tisdall J, Sukartono, Kusnarta IGM, McKenzie BM (2004) Physical properties of a clay loam soil mixed with sand. In 'SuperSoil 2004: 3rd Australian & New Zealand Soils Conference'. 5-9 December 2004. (The Regional Institute Ltd: Gosford, NSW) Available at: www.regional. org.au/au/asssi/supersoil2004/s14/poster/1585_gillj.htm

Grant CD, Groenevelt PH (2015) Weighting the differential water capacity to account for declining hydraulic conductivity in a drying coarse-textured soil. Soil Research 53, 386-391. doi: 10.1071/SR14258

Grant CD, Groenevelt PH, Robinson NI (2010) Application of the Groenevelt-Grant soil water retention model to predict the hydraulic conductivity. Soil Research 48, 447-458. doi: 10.1071/SR09198

Groenevelt PH, Grant CD, Semetsa S (2001) A new procedure to determine soil water availability. Australian Journal of Soil Research 39, 577-598. doi: 10.1071/SR99084

Hall D, Lemon J, Oliver Y, Gazey C, Davies S, Russell C, Witham N (2009) Managing south coast sandplain soils to yield potential. Bulletin No. 4773, October 2009. Department of Agriculture and Food, Western Australia. Albany, W. Aust.

Hall DJM, Jones HR, Crabtree WL, Daniels TL (2010) Claying and deep ripping can increase crop yields and profits on water repellent sands with marginal fertility in southern Western Australia. Australian Journal of Soil Research 48, 178-187. doi:10.1071/SR09078

Hamblin A, Richards Q, Blake J (1988) Crop growth across a toposequence controlled by depth of sand over clay. Australian Journal of Soil Research 26, 623-635. doi:10.1071/SR9880623

Hardie MA, Cotching WE, Doyle RB, Holz G, Lisson S, Mattem K (2011) Effect of antecedent soil moisture on preferential flow in a texture-contrast soil. Journal of Hydrology 398, 191-201. doi: 10.1016/j.jhydrol. 2010.12.008

Harper RJ, Gilkes RJ (2004) The effects of clay and sand additions on the strength of sandy topsoils. Australian Journal of Soil Research 42, 39-44. doi: 10.1071 /SR03063

Harper RJ, McKissock I, Gilkes RJ, Carter DJ, Blackwell PS (2000) A multivariate framework for interpreting the effects of soil properties, soil management and landuse on water repellency. Journal of Hydrology 231-232, 371-383. doi: 10.1016/S0022-1694(00)00209-2

Isbell R (2002) 'The Australian Soil Classification.' Revised edn. (CSIRO Publishing: Melbourne)

Kramer PJ (1949) 'Plant and soil water relationships.' (McGraw Hill Book Co.: New York)

Lipiec J, Walczak R, Witkowska-Walczak B, Nosalewicz A, Slowinska-Jurkiewicz A, Slawinski C (2007) The effect of aggregate size on water retention and pore structure of two silt loam soils of different genesis. Soil & Tillage Research 97, 239-246. doi: 10.1016/j.still.2007.10.001

MacGillivray JH, Doneen LD (1942) Soil moisture conditions as related to the irrigation of truck crops on mineral soils. Proceedings American Society for Horticultural Science 40, 483-492.

Marshall TJ, Holmes JW, Rose CW (1996) 'Soil physics.' (Cambridge University Press: Cambridge, UK)

Martinez E, Fuentes J-P, Silva P, Valle S, Acevedo E (2008) Soil physical properties and wheat root growth as affected by no-tillage and conventional tillage systems in a Mediterranean environment of Chile. Soil & Tillage Research 99, 232-244. doi: 10.1016/j.still.2008. 02.001

May R, South Australian Research and Development Institute, Rural Solutions SA, Grains Research and Development Corporation (Australia) (2006) 'Clay spreading and delving on Eyre Peninsula: a broadacre clay application manual for farmers, contractors and advisors.' (South Australian Research and Development Institute: Adelaide, S. Aust.)

Nang DN (2012) Plant availability of water in soils being reclaimed from the saline-sodic state. PhD Thesis, University of Adelaide, Adelaide, S. Aust.

National Committee on Soil and Terrain (2009) 'Australian soil and land survey field handbook.' 3rd edn. (CSIRO Publishing: Melbourne)

Or D, Wraith JM, Robinson DA, Jones SB (2012) Soil water content and water potential relationships. In 'Handbook of soil sciences: properties and processes'. 2nd edn. (Eds PM Huang, Y Li, ME Sumner) pp. 4-1-4-28. (CRC Press: Boca Raton, FL)

Parametric Technology Corporation (2007) 'Mathcad 14.' (Parametric Technology Company: Needham, MA, USA) Available at: www.ptc.com.

Rawls WJ, Brakensiek DL, Saxton KE (1982) Estimation of soil-water properties. Transactions of the American Society of Agricultural Engineers 25, 1316-1328.

Rayment GE (1992) 'Australian laboratory handbook of soil and water chemical methods.' (Eds GE Rayment, FR Higginson) (Inkata Press: Melbourne)

Rebbeck M, Lynch C, Hayman PT, Sadras VO (2007) Delving of sandy surfaced soils reduces frost damage in wheat crops. Australian Journal of Agricultural Research 58, 105-112. doi:10.1071/AR06097

Reynolds WD, Bowman BT, Brunke RR, Drury CF, Tan CS (2000) Comparison of tension infiltrometer, pressure infiltrometer, and soil core estimates of saturated hydraulic conductivity. Soil Science Society of America Journal 54, 1233-1241.

Rijtema PE (1969) Soil moisture forecasting. ICW Report No. 513. Instituut voor Cultuurtechniek en Waterhuishouding. University of Wageningen, Wageningen, The Netherlands.

Ritsema CJ, Dekker LW (2000) Preferential flow in water repellent sandy soils: principles and modeling implications. Journal of Hydrology 231-232, 308-319. doi:10.1016/S0022-1694(00)00203-1

Saxton KE, Rawls WJ (2006) Soil water characteristic estimates by texture and organic matter for hydrologic solutions. Soil Science Society of America Journal 70, 1569-1578. doi:10.2136/sssaj2005.0117

Shiel RS, Adey MA, Lodder M (1988) The effect of successive wet/dry cycles on aggregate size distribution in a clay texture soil. Journal of Soil Science 39, 71-80. doi:10.1111/j.1365-2389.1988.tb01195.x

Smith LH, Tiller KG (1977) A modified procedure for the more rapid determination of the clay content of soils. Divisional Report No. 19. CSIRO Division of Soils.

Tennant D, Scholz G, Dixon J, Purdie B (1992) Physical and chemical characteristics of duplex soils and their distribution in the south-west of Western Australia. Australian Journal of Experimental Agriculture 32, 827-843. doi: 10.1071/EA9920827

van Lier QJ, Metselaar K, van Dam JC (2006) Root water extraction and limiting soil hydraulic conditions estimated by numerical simulation. Vadose Zone Journal 5, 1264-1277. doi:10.2136/vzj2006.0056

Ward PR (1993) Generation of water repellence in sands, and its amelioration by clay addition. PhD Thesis, University of Adelaide, Adelaide, S. Aust.

WRB IWG (2007) 'World Reference Base for Soil Resources 2006. First update 2007.' (FAO: Rome) Available at: www.fao.org

http://dx.doi.org/10.1071/SR15115

Giacomo Betti (A), Cameron D. Grant (A,B), Robert S. Murray (A), and G. Jock Churchman (A)

(A) University of Adelaide, School of Agriculture, Food & Wine, Waite Campus PMB 1, Glen Osmond, SA 5064, Australia.

(B) Corresponding author. Email: cameron.grant@adelaide.edu.au

Table 1. Physical and chemical properties of bulk samples of the
A-horizon sand and the B-horizon clay-rich subsoil from Coonalpyn
and Bordertown, South Australia

Values in parentheses represent standard deviation

Soil horizon     Dry bulk density      % Clay             % Sand
                 (kg [m.sup.-3])    (<2 [micro]m)   (20-2000 [micro]m)

Coonalpyn, South Australia

A-horizon sand      1551 (13)             1                 97
B-horizon           1723 (21)            28                 71

Bordertown, South Australia

A-horizon sand      1499 (11)             2                 92
B-horizon           1806 (19)            47                 61

Soil horizon             PH           EC (dS [m.sup.-1])
                 (1:5 soil : water)   (1:5 soil : water)

Coonalpyn, South Australia

A-horizon sand          6.2                  0.02
B-horizon               7.1                  0.04

Bordertown, South Australia

A-horizon sand          6.5                  0.02
B-horizon               8.4                  0.18

Table 2. Clay content, texture, measured and predicted bulk density,
hydraulic conductivity and soil resistance for sand mixed with 0-100%
clay-rich subsoil in different aggregate sizes (<2, 6,20 and 45 mm)

Values in parentheses represent standard deviation for the
[[rho].sub.b] measured data and coefficient of variation for the
[K.sub.s] data, n.a., Not applicable

Clay-rich   Nominal diam.     % <2     Textural   [[rho].sub.b]
subsoil       clay-rich     [micro]m    class     (kg [m.sup.-3])
(% by wt)    aggregates
               (mm)                               Measured    Predicted
                                                                from
                                                                Eqn 6

Coonalpyn

0%              n.a.           1         Sand     1551 (13)     n.a.

10%              <2            4         Sand     1630 (10)     1630
                  6                               1640 (10)
                 20                               1650 (20)

20%              <2            7         Sand     1610 (5)      1640
                  6                               1630 (10)
                 20                               1640 (10)
                 45                               1660 (10)

40%              <2            12       Loamy     1590 (10)     1660
                  6                      sand     1620 (10)
                 20                               1670 (10)
                 45                               1680 (20)

60%              <2            18       Sandy     1570 (10)     1680
                  6                      loam     1600 (10)
                 20                               1680 (30)

100%          0< x <45         28       Sandy     1723 (21)     n.a.
                                         clay
                                         loam

Bordertown

0%              n.a.           2         Sand     1499 (10)     n.a.

10%              <2            7         Sand     1530 (10)     1517
                  6                               1510 (5)
                 20                               1520 (20)

20%              <2            11       Loamy     1510 (20)     1545
                  5                      sand     1520 (10)
                 20                               1520 (20)

40%              <2            21       Sandy     1530 (30)     1603
                  6                      clay     1520 (20)
                 20                      loam     1580 (20)

60%              <2            30       Sandy     1500 (10)     1667
                  6                      clay     1500 (40)
                 20                      loam     1620 (10)

100%          0< x <45         47        clay     1806 (19)     n.a.

Clay-rich   Nominal diam.   [K.sub.s] (m [s.sup.-1])
subsoil     clay-rich
(% by wt)   aggregates
            (mm)

Coonalpyn

0%              n.a.        1.3 x [10.sup.-4] (0.14)

10%              <2         7.7 x [10.sup.-5] (0.05)
                  6         7.9 x [10.sup.-5] (0.16)
                 20         1.1 x [10.sup.-4] (0.04)

20%              <2         3.2 x [10.sup.-5] (0.03)
                  6         7.0 x [10.sup.-5] (0.01)
                 20         9.4 x [10.sup.-5] (0.09)
                 45         9.1 x [10.sup.-5] (0.08)

40%              <2         4.7 x [10.sup.-5] (0.54)
                  6         3.3 x [10.sup.-5] (0.16)
                 20         5.0 x [10.sup.-5] (0.13)
                 45         7.5 x [10.sup.-5] (0.02)

60%              <2         3.5 x [10.sup.-6] (0.01)
                  6         1.2 x [10.sup.-5] (0.04)
                 20         5.0 x [10.sup.-5] (0.16)

100%          0< x <45      4.3 x [10.sup.-8] (0.16)

Bordertown

0%              n.a.        5.2 x [10.sup.-5] (0.10)

10%              <2         1.6 x [10.sup.-5] (0.27)
                  6         3.0 x [10.sup.-5] (0.05)
                 20         3.0 x [10.sup.-5] (0.13)

20%              <2         5.3 x [10.sup.-6] (0.20)
                  5         2.4 x [10.sup.-5] (0.11)
                 20         2.5 x [10.sup.-5] (0.06)

40%              <2         3.1 x [10.sup.-7] (0.34)
                  6         9.4 x [10.sup.-6] (0.17)
                 20         1.4 x [10.sup.-5] (0.05)

60%              <2         1.7 x [10.sup.-7] (1.00)
                  6         6.6 x [10.sup.-6] (0.20)
                 20         8.6 x [10.sup.-6] (0.04)

100%          0< x <45      7.1 x [10.sup.-9] (0.19)

Clay-rich   Nominal diam.         Rijtema K(h) function
subsoil     clay-rich            parameters (Eqns 12, 13)
(% by wt)   aggregates
            (mm)            Wet-end [alpha]    Dry-end [alpha]
                             ([m.sup.-1])       ([m.sup.2,4]
                                                [day.sup.-1])

Coonalpyn

0%              n.a.             13.8         9.98 x [10.sup.-6]

10%              <2              13.80        1.73 x [10.sup.-4]
                  6              13.80
                 20              13.80

20%              <2              8.22         1.73 x [10.sup.-4]
                  6              13.80
                 20              13.80
                 45              13.80

40%              <2              5.00         8.34 x [10.sup.-5]
                  6              8.22
                 20              13.80
                 45              13.80

60%              <2              82.20        3.58 x [10.sup.-4]
                  6              82.20
                 20              8.22

100%          0< x <45           3.53         5.71 x [10.sup.-5]

Bordertown

0%              n.a.             13.80        9.98 x [10.sup.-6]

10%              <2              8.22         1.73 x [10.sup.-4]
                  6              8.22
                 20              8.22

20%              <2              5.00         2.6 x [10.sup.-4]
                  5              8.22
                 20              8.22

40%              <2              5.62         4.2 x [10.sup.-4]
                  6              5.00
                 20              8.22

60%              <2              2.69         3.58 x [10.sup.-4]
                  6              5.62         3.58 x [10.sup.-4]
                 20              5.00         5.33 x [10.sup.-4]

100%          0< x <45           2.48         2.68 x [10.sup.-5]

Clay-rich   Nominal diam.        Parameter in SR(h) =
subsoil     clay-rich             ([sigma][h.sup.b]
(% by wt)   aggregates                 (Eqn 10)
            (mm)
                            [alpha] (equal    b     SR(150)
                              to SR(1))              (MPa)

Coonalpyn

0%              n.a.             0.42        0.02    0.46

10%              <2              0.45        0.15    0.94
                  6              0.43        0.07    0.60
                 20              0.44        0.01    0.48

20%              <2              0.48        0.28    1.94
                  6              0.48        0.05    0.61
                 20              0.40        0.07    0.58
                 45              0.46        0.17    1.01

40%              <2              0.35        0.37    2.26
                  6              0.31        0.14    0.63
                 20              0.36        0.11    0.62
                 45              0.54        0.22    1.66

60%              <2              0.70        0.31    3.28
                  6              0.33        0.25    1.16
                 20              0.37        0.10    0.61

100%          0< x <45           0.39        0.54    5.80

Bordertown

0%              n.a.             0.47        0.03    0.55

10%              <2              0.53        0.17    1.22
                  6              0.43        0.14    0.87
                 20              0.44        0.06    0.6

20%              <2              0.42        0.35    2.40
                  5              0.43        0.20    1.18
                 20              0.33        0.08    0.50

40%              <2              0.32        0.41    2.56
                  6              0.21        0.36    1.3
                 20              0.18        0.38    1.25

60%              <2              0.23        0.48    2.52
                  6              0.22        0.39    1.60
                 20              0.06        0.82    3.65

100%          0< x <45           0.32        0.60    6.45

Table 3. Minimum and maximum matric heads ([h.sub.min] and
[h.sub.max], m) used as limits in Eqns 9, 10 and 16 to calculate the
integral water capacity (IWC) for mixtures of sand and clay-rich
subsoil

Values in parentheses represent standard deviation, n.a., Not
applicable

Clay-rich   Nominal diam.          Poor soil aeration Eqn 9
subsoil       clay-rich
(% by wt)    aggregates        [h.sub.minA]         [h.sub.maxA]
                (mm)        [[theta].sub.minA]   [[theta].sub.maxA]
                              0.1 [m.sup.3]        0.15 [m.sup.3]
                              air [m.sup.-3]       air [m.sup.-3]

Coonalpyn

0%              n.a.           0.23 (0.00)          0.29 (0.00)

10%               2            0.31 (0.00)          0.39 (0.01)
                  6            0.33 (0.00)          0.44 (0.04)
                 20            0.30 (0.01)          0.37 (0.01)

20%               2            0.27 (0.03)          0.46 (0.04)
                  6            0.18 (0.01)          0.29 (0.02)
                 20            0.21 (0.01)          0.32 (0.01)
                 45            0.20 (0.02)          0.30 (0.02)

40%               2            0.37 (0.02)          0.80 (0.05)
                  6            0.26 (0.01)          0.46 (0.02)
                 20            0.24 (0.01)          0.38 (0.02)
                 45            0.24 (0.00)          0.36 (0.00)

60%               2            0.29 (0.01)          0.81 (0.03)
                  6            0.11 (0.00)          0.28 (0.02)
                 20            0.17 (0.01)          0.33 (0.03)

100%            n.a.            4.3 (0.81)          21.5 (24.3)

Bordertown

0%              n.a.           0.32 (0.00)          0.41 (0.00)

10%               2            0.34 (0.04)          0.44 (0.04)
                  6            0.37 (0.01)          0.47 (0.00)
                 20            0.37 (0.02)          0.48 (0.01)

20%               2            0.28 (0.06)          0.46 (0.08)
                  6            0.39 (0.04)          0.54 (0.05)
                 20            0.41 (0.01)          0.54 (0.01)

40%               2            0.35 (0.01)          0.58 (0.03)
                  6            0.47 (0.10)          0.93 (0.20)
                 20            0.25 (0.08)          0.43 (0.11)

60%               2            0.36 (0.09)          0.65 (0.15)
                  6            0.46 (0.07)          1.11 (0.18)
                 20            0.21 (0.03)          0.47 (0.08)

100%            n.a.            9.8 (6.3)           44.5 (29.7)

Clay-rich   Nominal diam.         Soil resistance Eqn 10
subsoil       clay-rich
(% by wt)    aggregates     [h.sub.minSR] at   [h.sub.maxSR] at
                (mm)          SR = 0.5 MPa       SR = 2.5 MPa

Coonalpyn

0%              n.a.          >[10.sup.3]        >[10.sup.3]

10%               2               2.0            >[10.sup.3]
                  6               10.9           >[10.sup.3]
                 20           >[10.sup.3]        >[10.sup.3]

20%               2               1.2                376
                  6               2.1            >[10.sup.3]
                 20               20.1           >[10.sup.3]
                 45               1.6            >[10.sup.3]

40%               2               2.6                196
                  6               28.9           >[10.sup.3]
                 20               18.9           >[10.sup.3]
                 45               0.70               957

60%               2               0.3                 54
                  6               5.0            >[10.sup.3]
                 20               19.4           >[10.sup.3]

100%            n.a.              1.6                 32


Bordertown

0%              n.a.              6.5            >[10.sup.3]

10%               2               0.68           >[10.sup.3]
                  6               2.8            >[10.sup.3]
                 20               8.7            >[10.sup.3]

20%               2               1.7                169
                  6               2.1            >[10.sup.3]
                 20               150            >[10.sup.3]

40%               2               2.9                142
                  6               10.6               923
                 20               13.7               946

60%               2               5.2                146
                  6               7.8                469
                 20               13.7                95

100%            n.a.              2.1                 31

Clay-rich   Nominal diam.    Hydraulic conductivity Eqn
subsoil       clay-rich
(% by wt)    aggregates     [h.sub.limK] at [M.sub.lim] =
                (mm)         0.1 [cm.sup.2] [day.sup.-1]

Coonalpyn

0%              n.a.                     4.8

10%               2
                  6                      101
                 20

20%               2
                  6                      101
                 20
                 45

40%               2
                  6                      70
                 20
                 45

60%               2
                  6                      123
                 20

100%            n.a.                     53

Bordertown

0%              n.a.                     4.8

10%               2
                  6                      101
                 20

20%               2
                  6                      115
                 20

40%               2
                  6                      127
                 20

60%               2                      123
                  6                      123
                 20                      131

100%            n.a.                     23

Table 4. Parameters from Eqn 6 to describe the water-retention
curves, plus estimates of plant-available water (PAW, mm [m.sup.-1])
from Eqns 4 and 5 for mixtures of sand and clay-rich subsoil, as well
as for similar textures from the literature

Values in parentheses represent standard deviation. SSE represents
the sum of square errors between the predicted and measured
volumetric water contents

Site         % Clay-rich    Size of         Parameters for Eqn 6
             aggregates    clay-rich
             in mix        aggregates   [k.sub.0]   [k.sub.1]     n
                              (mm)         (m)

Coonalpyn          0%       n.a.          0.276       0.347     1.584

                  10%          2          0.347       0.334     1.789
                               6          0.366       0.322     1.549
                              20          0.331       0.332     1.892

                  20%          2          0.305       0.302     0.873
                               6          0.269       0.315     1.040
                              20          0.246       0.326     0.974
                              45          0.245       0.316     1.053

                  40%          2          0.403       0.289     0.638
                               6          0.297       0.318     0.737
                              20          0.286       0.315     0.980
                              45          0.275       0.317     1.149

                  60%          2          0.388       0.305     0.446
                               6          0.170       0.341     0.432
                              20          0.209       0.321     0.641

                  100%      n.a.          9.534       0.344     0.248

Bordertown         0%       n.a.          0.367       0.313     1.67

                  10%          2          0.392       0.306     1.696
                               6          0.410       0.314     1.947
                              20          0.411       0.308     1.724

                  20%          2          0.349       0.299     0.966
                               6          0.409       0.288     1.461
                              20          0.440       0.295     1.755

                  40%          2          0.471       0.328     0.816
                               6          0.612       0.312     0.638
                              20          0.341       0.316     0.788

                  60%          2          0.546       0.337     0.672
                               6          0.687       0.304     0.514
                              20          0.275       0.310     0.544

                  100%      n.a.          0.373       0.298     1.696

Site         % Clay-rich    Size of         Parameters for Eqn 6
             aggregates    clay-rich
             in mix        aggregates   [[theta].          SSE
                              (mm)       sub.a]

Coonalpyn          0%       n.a.          0.040     4.2 x [10.sup.-4]

                  10%          2          0.051     1.2 x [10.sup.-3]
                               6          0.052     3.1 x [10.sup.-3]
                              20          0.052     2.2 x [10.sup.-3]

                  20%          2          0.057     1.6 x [10.sup.-4]
                               6          0.058     1.9 x [10.sup.-3]
                              20          0.058     1.4 x [10.sup.-3]
                              45          0.059     2.1 x [10.sup.-3]

                  40%          2          0.113     3.3 x [l0.sup.-4]
                               6          0.115     1.6 x [10.sup.-3]
                              20          0.119     2.0 x [10.sup.-3]
                              45          0.119     9.7 x [10.sup.-4]

                  60%          2          0.146     2.6 x [10.sup.-4]
                               6          0.149     1.1 x [10.sup.-3]
                              20          0.157     1.5 x [10.sup.-4]

                  100%      n.a.          0.193     10.0 x [10.sup.-4]

Bordertown         0%       n.a.          0.043     5.4 x [10.sup.-4]

                  10%          2          0.050     6.9 x [10.sup.-4]
                               6          0.049     6.4 x [10.sup.-4]
                              20          0.050     9.5 x [10.sup.-4]

                  20%          2          0.082     1.5 x [10.sup.-3]
                               6          0.081     7.9 x [10.sup.-4]
                              20          0.081     8.2 x [10.sup.-4]

                  40%          2          0.103     1.2 x [10.sup.-3]
                               6          0.102     1.2 x [10.sup.-3]
                              20          0.106     1.6 x [10.sup.-3]

                  60%          2          0.162     1.3 x [10.sup.-3]
                               6          0.163     1.3 x [10.sup.-3]
                              20          0.175     1.4 x [10.sup.-3]

                  100%      n.a.          0.250     9.5 x [10.sup.-5]

Site         % Clay-rich    Size of      Estimates of Plant
             aggregates    clay-rich    Available Water (PAW,
             in mix        aggregates      mm [m.sup.-1])
                              (mm)
                                          Eqn 4      Eqn 5

Coonalpyn          0%       n.a.         40 (0)      n.a.

                  10%          2         47 (4)      50 (1)
                               6         46 (6)
                              20         39 (1)

                  20%          2         89 (4)      59 (2)
                               6         71 (2)
                              20         73 (4)
                              45         64 (3)

                  40%          2        118 (10)     79 (5)
                               6        103 (5)
                              20         79 (6)
                              45         65 (2)

                  60%          2        126 (2)     100 (7)
                               6        109 (2)
                              20         94 (4)

                  100%      n.a.        145 (13)     n.a.

Bordertown         0%       n.a.         53 (0)      n.a.

                  10%          2         57 (6)      62 (1)
                               6         51 (2)
                              20         60 (5)

                  20%          2         90 (12)     71 (2)
                               6         69 (8)
                              20         62 (5)

                  40%          2        134 (6)      91 (5)
                               6        152 (4)
                              20        107 (14)

                  60%          2        155 (22)    113 (8)
                               6        152 (4)
                              20        110 (15)

                  100%      n.a.        161 (14)     n.a.

Site         % Clay-rich    Size of     Estimates of Plant Available
             aggregates    clay-rich     Water (PAW, mm [m.sup.-1])
             in mix        aggregates
                              (mm)
                                        Based on soil texture   Source

Coonalpyn          0%       n.a.           Sand      13 to      1, 3
                                                       80
                  10%          2
                               6
                              20

                  20%          2
                               6
                              20
                              45

                  40%          2        Loamy sand     80        4
                               6
                              20
                              45

                  60%          2        Sandy loam   48 to      2, 3, 4
                               6                      188
                              20

                  100%      n.a.        Sandy clay   59 to       1, 2
                                           loam       236

Bordertown         0%       n.a.

                  10%          2           Sand      13 to       1, 3
                               6                       80
                              20

                  20%          2        Loamy sand     80          4
                               6
                              20

                  40%          2        Sandy clay   59 to       1, 2
                               6           loam       236
                              20

                  60%          2
                               6
                              20

                  100%      n.a.         Clay      150 to        1, 4
                                                      236

Sources: MacGillivray and Doneen (1942); Kramer (1949, table 3, p. 61);
Marshall et al. (1996, table 10.1, p. 251); and Or et al. (2012, drawn
from fig. 4.8).

Table 5. Estimates of integral water capacity (IWC, mm [m.sup.-1])
for the individual and combined limiting factors for mixtures of sand
and clay-rich subsoil

Values in parentheses represent standard deviation, n.a., not
applicable. For mixtures containing aggregates of 20 or 45 mm
diameter, IWC was estimated using Eqn 18

Clay-rich    Nominal diameter    [[integral].sup.
subsoil (%      clay-rich       [infinity].sub.0]
by weight)   aggregates (mm)          C(h)dh

Coonalpyn

0%                 n.a.              347 (0)

10%                 2                334 (9)
                    6                317 (4)
              [greater than            348
             or equal to] 20

20%                 2                303 (11)
                    6                329 (2)
              [greater than            349
             or equal to] 20

40%                 2                289 (13)
                    6                318 (9)
              [greater than            350
             or equal to] 20

60%                 2                305 (2)
                    6                341 (8)
              [greater than            352
             or equal to] 20

100%               n.a.              355 (37)

Bordertown

0%                 n.a.              313 (0)

10%                 2                306 (4)
                    6                314 (4)
              [greater than          312 (8)
             or equal to] 20

20%                 2                299 (2)
                    6                288 (59)
              [greater than          312 (14)
             or equal to] 20

40%                 2               328 (11)
                    6                312 (6)
              [greater than          310 (30)
             or equal to] 20

60%                 2                337 (7)
                    6                304 (14)
              [greater than          308 (47)
             or equal to] 20

100%               n.a.              305 (85)

Clay-rich    Nominal diameter   IWC from Eqns 8 and 18 accounting for
subsoil (%      clay-rich         all or individual limiting factors
by weight)   aggregates (mm)

                                  All       Aeration        Soil
                                factors    factor only   resistance
                                                         factor only

Coonalpyn

0%                 n.a.          46 (0)      228 (0)       347 (0)

10%                 2            80 (4)      208 (9)       334 (9)
                    6            79 (7)      194 (5)       317 (4)
              [greater than        71          219           348
             or equal to] 20

20%                 2            90 (3)     178 (11)      292 (10)
                    6            97 (7)      211 (4)       329 (2)
              [greater than        96          216           349
             or equal to] 20

40%                 2           116 (8)     162 (11)       269 (8)
                    6           112 (4)     189 (10)       317 (9)
              [greater than       146          208           350
             or equal to] 20

60%                 2           103 (2)      182 (3)       226 (2)
                    6           110 (2)      214 (9)      323 (11)
              [greater than       135          136           352
             or equal to] 20

100%               n.a.          13 (3)     224 (36)       133 (4)

Bordertown

0%                 n.a.          64 (0)      200 (0)       313 (0)

10%                 2            69 (6)      196 (4)       303 (4)
                    6            67 (2)      194 (0)       314 (4)
              [greater than      82 (4)      185 (4)       312 (8)
             or equal to] 20

20%                 2           103 (13)     186 (0)       291 (2)
                    6            82 (8)      165 (1)       287 (6)
              [greater than     100 (9)      181 (8)      312 (14)
             or equal to] 20

40%                 2           164 (7)     213 (12)      314 (10)
                    6           159 (6)     193 (12)       301 (6)
              [greater than     137 (16)    171 (16)      310 (30)
             or equal to] 20

60%                 2           170 (16)     223 (4)       316 (4)
                    6           148 (5)     190 (13)      278 (13)
              [greater than     105 (16)    105 (16)      308 (47)
             or equal to] 20

100%               n.a.         24 (39)     180 (85)      126 (59)

Clay-rich    Nominal diameter   IWC from Eqns 8 and 18 accounting
subsoil (%      clay-rich         for all or individual limiting
by weight)   aggregates (mm)                 factors

                                K[(h).sub.WET]    K[(h).sub.DRY]
                                  factor only       factor only

Coonalpyn

0%                 n.a.             48 (0)            345 (0)

10%                 2               81 (5)            334 (9)
                    6               80 (8)            301 (8)
              [greater than           76                347
             or equal to] 20

20%                 2               100 (4)           100 (4)
                    6               99 (8)            99 (8)
              [greater than           104               347
             or equal to] 20

40%                 2              141 (13)          141 (13)
                    6               116 (6)           116 (6)
              [greater than           163               349
             or equal to] 20

60%                 2               304 (2)           304 (2)
                    6               135 (3)           135 (3)
              [greater than           225               351
             or equal to] 20

100%               n.a.            301 (39)           194 (5)

Bordertown

0%                 n.a.             66 (0)            311 (0)

10%                 2               73 (7)            306 (4)
                    6               67 (3)            314 (4)
              [greater than         87 (7)            311 (7)
             or equal to] 20

20%                 2              113 (14)           298 (0)
                    6               83 (9)            288 (6)
              [greater than        107 (15)          310 (15)
             or equal to] 20

40%                 2               187 (8)          325 (10)
                    6              180 (10)           303 (6)
              [greater than        151 (31)          309 (31)
             or equal to] 20

60%                 2              208 (23)           329 (7)
                    6               183 (4)          284 (13)
              [greater than        198 (47)          308 (48)
             or equal to] 20

100%               n.a.            271 (48)          177 (718)
COPYRIGHT 2016 CSIRO Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Betti, Giacomo; Grant, Cameron D.; Murray, Robert S.; Churchman, G. Jock
Publication:Soil Research
Article Type:Report
Date:May 1, 2016
Words:11077
Previous Article:Biochemical properties of highly mineralised and infertile soil modified by acacia and spinifex plants in Northwest Queensland, Australia.
Next Article:Application of neural network and grey relational analysis in ranking the factors affecting runoff and sediment yield under simulated rainfall.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters