Changes in soils irrigated with saline groundwater containing excess bicarbonate.
Subject: Soils, Salts in (Research)
Groundwater flow (Research)
Author: Gardner, W.K.
Pub Date: 12/01/2004
Publication: Name: Australian Journal of Soil Research Publisher: CSIRO Publishing Audience: Academic Format: Magazine/Journal Subject: Agricultural industry; Earth sciences Copyright: COPYRIGHT 2004 CSIRO Publishing ISSN: 0004-9573
Issue: Date: Dec, 2004 Source Volume: 42 Source Issue: 7
Topic: Event Code: 310 Science & research
Accession Number: 126317205
Full Text: Changes in soil properties caused by irrigation with saline groundwater (approx. 2 dS/m) containing

excess bicarbonate were measured on Vertosols and Sodosols in the West Wimmera, Victoria, Australia. Irrigation

caused soil pH to increase, and where this had risen sufficiently (approx. 8.0), the sodium absorption ratio (SAR)

of 1 : 5 soil extracts also increased, presumably due to precipitation of calcium and magnesium carbonates. Salt

only accumulated when the SAR of 1 : 5 soil extracts was high. In contrast to previous studies, SAR of the soil

extracts was not correlated with exchangeable sodium percentage (ESP) of the exchange complex, nor with soil pH.

SAR values rose with irrigation once pH exceeded 8, suggesting that carbonate formation was incomplete due to

insufficient bicarbonate. The results imply that gypsum application may ameliorate soil properties even if amounts

applied are not sufficient to alter ESP.

Introduction

'Hostile' subsoils are seen as major barriers to crop production on many soils in southern Australia. Salinity, sodicity, and boron toxicity have been identified as the likely soil characteristics associated with low root growth and crop production. Research has not yet identified exactly how these characteristics interact to affect plants.

O'Leary et al. (2003) used electromagnetic sensing to measure soil boron, chloride, and water content, but obtained only low correlations with crop yields. Nuttall et al. (2003) found only moderate correlations between crop yields and electrical conductivity (EC), exchangeable sodium percentage (ESP), and soluble boron levels in subsoils. Armstrong and co-workers found in a glasshouse study that boron tolerance did not satisfactorily explain root growth in subsoils, and postulated that some other mechanism was operative (Armstrong et al. 2003).

ESP is used to assess sodicity of soil and is defined as the percentage of the cation exchange complex occupied by sodium, while sodium absorption ratio (SAR) is used to assess sodicity of irrigation waters and soil solutions and is defined as the concentration of sodium divided by the square root of the sum of calcium plus magnesium, where concentrations are in mmol/L. Criteria for evaluating soil sodicity, and their relationships to soil properties, have been reviewed by Rengasamy and Churchman (1999). They reported that the SAR of the soil solution was related to the ESP of the exchange complex, but that several factors affected this relationship and it should only be used with caution. Gupta and Abrol (1990) also reported that SAR and ESP were related.

The work presented in this paper was part of a study to investigate the effect of centre pivot irrigation with saline groundwater (approx. 2 dS/m) containing excess bicarbonate on soil properties of Vertosols and Sodosols in the West Wimmera. This indicated that consideration of SAR improves soil sodicity assessment.

Materials and methods

Location and sampling of West Wimmera Pivots

In 2001, a study was undertaken of 9 centre pivot irrigation installations in the Neuarpurr groundwater supply protection area, between Edenhope and the Victoria-South Australia border (E500800-516500 and N5907600-5940300 UTM 1984). Soil types were Vertosols and Sodosols (a detailed description of soil types and land systems is given in Baxter et al. 1998). Two areas were examined at pivot 3, four at pivot 4, and one at each of the other pivots. At each sampling area, comparisons were made under the pivot approximately 50-75 m from the perimeter and outside the pivot (at a similar distance) by taking five 50-mm soil cores to 1 m depth using a hydraulic push sampler. The soil cores were taken in a line at 10 m intervals. Soil cores were divided into 20-cm depth increments and bulked.

Centre pivot 3 was installed in 1989 and has grown 7 white clover (Trifolium repens) seed crops, 3 shaftal clover (17 resupinatum) hay crops, 1 canola, and 1 spring sown barley crop. Two locations were sampled, one a loam over ferric nodules over clay area (Ferric Sodosol), the other a heavy clay area (Vertosol). The farmer stated that the heavy clay area has never performed well, often becoming waterlogged in winter. Gypsum had been applied at 2.5 t/ha in 1989, 1990, and 1995 and at 3.7 t/ha in 1996. In 2000, 3 t/ha was applied to the Vertosol area and 1.25 t/ha to the Sodosol area. Water metering has recently been installed on some pivots; otherwise, it was estimated that 400 mm of irrigation water was applied to white clover crops, and 250 mm to other crop types.

Centre pivot 4 was installed in 1983 and had produced 13 white clover seed crops, 4 wheat crops, and 1 oaten hay crop. Gypsum at 2.5 t/ha had been applied in 1984, 1990, 1994, and 1999. The pivot was sampled at 4 separate places, which varied as follows:

(i) location 1--loam over clay (Solodised Solonetz),

(ii) location 2--loam over a ferruginised (iron) nodule layer over clay (Ferric Sodosol),

(iii) location 3--light clay grading to heavy clay (Vertosol),

(iv) location 4--clay (Vertosol).

Soil samples were sent to Pivot laboratories (NATA-accredited) and measurements made on 1 : 5 soil water extracts of pH(water), EC (1 : 5), and sodium, potassium, calcium, and magnesium concentrations. Chloride concentration was also measured. Exchangeable cations were measured using 1 M ammonium chloride at pH 8.5 with a prewash to minimise the effects of sparingly soluble salts (Rayment and Higginson 1992).

Samples were taken of irrigation waters by pumping for at least 30 min, and then collecting 1 L of water into a plastic container filled completely and sealed securely. Salt contents, pH, bicarbonate, and carbonate content were measured.

Estimation of soil salinity levels

Using standard conversion factors for soil texture, the EC 1 : 5 values can be used to estimate the EC of the soil solution when the soil is at field capacity (E[C.sub.sat]). It is then possible to estimate salt effects on plant growth using published response curves (Maas and Hoffman 1977; Maas 1990)

This conversion is subject to some errors due to the effects of sparingly soluble salts (overestimation) and formation of non conducting ion pairs at high pH (underestimation), but these errors should not be sufficient to negate the overall trends. Currently, no Australian laboratories offer direct measurement of saturated extracts.

An overall measure of soil salinity at each location was calculated as a weighted average, which takes into account greater water use in the upper soil levels. A typical calculation is shown (Table I), using the weighting values typical of a soil where water is extracted to 1 m depth (Shaw 1999).

Sodicity of the soil solution

The SAR was calculated on 1 : 5 extracts.

Results

Analysis of irrigation waters

The irrigation groundwaters were all slightly alkaline (Table 2) with EC varying from 1.4 to 2.7dS/m, and contained significant amounts of bicarbonate (in excess of calcium plus magnesium, 1.3-6.2 mM/L)

The Victorian Irrigation Research and Advisory Services Committee (1980) classify such waters as significantly saline, requiring good drainage if salt accumulation is to be avoided with irrigation. The excess of bicarbonate over calcium plus magnesium indicates that its application will cause an alkaline reaction in the soil.

Soils analyses

Pivot 4

Weighted average salt contents. Some salt accumulation occurred with irrigation at all 4 locations, but it was particularly severe at location 4, resulting in a 25% potential yield loss (Table 3).

Soil pH. Soil pH in the surface layers increased with irrigation (Fig. 1). The soil at all locations was very alkaline at depth; however, location 4 had the highest surface pH outside the pivot, and had become more alkaline than the other locations, reaching nearly pH 8 at the surface under irrigation.

Sodicity of the soil solution. The SAR rose with irrigation particularly at location 4, but some increase was evident at 80 cm depth at location 1 (Fig. 2).

The same amount of salt was applied in the irrigation water to all the locations examined since irrigation began in 1983: however, salt mainly accumulated at location 4. Location 4 was initially more alkaline, and addition of bicarbonate in the irrigation water caused calcium and magnesium in the soil solution to precipitate as carbonates (hence the increased SAR values). This in turn may be expected to cause lower soil permeability and therefore result in accumulation of salt. The other locations were initially more acid, and the bicarbonate in the irrigation water resulted in higher pH values, with little change in SAR except at the 80 cm depth of location 1, and salt accumulation did not occur.

Pivot 3

Weighted average salt contents. Salt accumulation and potential yield loss in white clover was worse at location 2, resulting in a 54% potential yield loss under the pivot, an increase of 29% compared with the area outside the pivot (Table 4).

Soil pH. The pH has increased with irrigation at both locations; however, location 2 was initially more alkaline than location 1, and the pH in the top 40 cm increased to higher values with irrigation (Fig. 3).

Sodicity of the soil solution. The SAR was high at depth at location 2, both inside and outside the pivot, and increased at all depths with irrigation. At location 1, SAR increased below 80cm depth with irrigation, and was low at all depths outside the pivot (Fig. 4).

SAR at location 2 was high at depth outside the pivot, and salt content even in the absence of irrigation was already sufficiently high to cause significant potential yield reduction in white clover. This suggests that salts falling as rainfall have accumulated in the profile to a greater extent at this location, presumably due to the lower soil permeability bought about by the high SAR. It might also reflect differences in geology or history of location 2, but its proximity to location 1 makes this unlikely. Irrigation has further increased the SAR values at all depths at location 2, and salts have accumulated as irrigation has occurred.

SAR at Location 1 was low outside the pivot, but addition of bicarbonate in the irrigation water caused calcium and magnesium in the soil solution to precipitate as carbonates (thus the increased SAR values). This in turn resulted in lower soil permeability and salt accumulation.

Aggregated data (all pivots)

Chloride content and soil solution sodicity. Chloride content of the soil is a measure of salt accumulation as it is not subject to errors likely with EC values of 1 : 5 extracts (see above). Chloride levels tended to increase as SAR increased (Fig. 5), indicating a general relationship between these 2 parameters. The scatter evident in the graph at SAR values >15 may be due to equilibrium salt contents not yet being reached. As more salts are added over time, it is likely that chloride concentrations will increase until they lie closer to the relationship between SAR and chloride evident at lower SAR values.

pH and SAR. High values of SAR only occurred at high pH (>8.0); however, low SAR occurred at all pH values (Fig. 6). It is to be expected that high SAR only occurred at high pH, as the mechanism for calcium and magnesium removal from the soil solution is most likely carbonate precipitation, which only occurs at pH >8.0. The occurrence, however, of low SAR at high pH suggests that while carbonate precipitation was possible in these amples, it did not completely remove the calcium and magnesium from the soil solution, probably because there was insufficient carbonate in the system to allow the reactions to proceed to completion. In other words, these soils have not reached equilibrium with regard to carbonate formation due to insufficient reactants. This is supported by the rise in SAR measured as bicarbonate was added in irrigation water. Thus, SAR of the soil solution cannot be predicted simply from considerations of pH.

Exchangeable sodium percentage and SAR. Two relationships were evident, one group of points in which ESP varied from approximately 3 to 35% while SAR remained <5, and a second group in which ESP varied from 15 to 35 while SAR values were high (5-37) (Fig. 7). Overall, it is not possible to predict SAR values from consideration of ESP unless an explanation is found for the 2 relationships evident in the data. The incomplete precipitation of carbonates could explain this difference.

Discussion

The observations reported in this study strongly suggest that soil solution sodicity (as measured by SAR on a 1 : 5 extract) is an important soil parameter with predictive ability for plant performance. There are 3 possibilities to explain this relationship.

Soil permeability

The data from the centre pivots showed a relationship between salt accumulation, chloride content, and SAR. If differences due to soil genesis are ignored (possible given the close proximity of the sites and the similar geology), the salt contents of the soils outside the pivots are the equilibrium between salt added in rain and removed by leaching (dependent on soil permeability). Inside the pivots, salt accumulation results from the balance between salt addition in applied water and removal in deep drainage, again a function of soil permeability.

The relationship between salt content and SAR of the soils therefore indicates that soil permeability is influenced by SAR. Gupta and Abrol (1990) also reported that soil permeability was reduced as SAR increased. Plant growth may be affected directly due to insufficient voids to allow root penetration in areas of lower permeability.

Interaction with salt effects

Plant response to salt is markedly affected by the cations present in the soil solution, in this case estimated by the SAR value. Lauchli and Epstein (1990) reported that osmotically similar salt solutions caused markedly greater reductions in plant growth as the SAR of the solution increased.

Specific ion effects

In addition to osmotic effects, high levels of sodium can directly affect plant nutrition, primarily as induced calcium and magnesium deficiencies at high SAR values. Root growth has been shown to be particularly sensitive to calcium deficiency (Snowball and Robson 1983).

It is therefore proposed that SAR is a useful addition to current soil parameters used to explain plant response in 'hostile' subsoils, it incorporates effects on soil permeability (which may directly affect the ability of roots to penetrate the soil), modifies plant susceptibility to damage from any specified osmotic salt stress, and indicates the likelihood of calcium and magnesium deficiencies induced by a relative excess of sodium ions.

Gupta and Abrol (1990) and Rengasamy and Churchman (1999) reported that a relationship exists between ESP and SAR; however, this was not evident in the data presented here. In this instance, 2 relationships are apparent, and it is suggested that incomplete precipitation of magnesium and calcium in the soil solution to carbonates may underlie this. This assertion is supported by the increase in SAR as bicarbonate was added in the irrigation waters, and by the occurrence of a range of SAR values at pH levels at which carbonate formation is possible.

The relationship between SAR and ESP reported by Gupta and Abrol (1990) and Rengasamy and Churchman (1999) was determined on SAR values of 'equilibrated soil solutions'. It is not clear from these articles whether the method of preparing an 'equilibrated soil solution' affects the outcome, although Rengasamy and Churchman (1999) urge caution on this matter. In this study, SAR was determined on 1:5 extracts, as no Australian laboratory currently offers alternatives such as saturation extracts. SAR values determined at a 1:5 dilutions will vary from that determined on saturated extracts due to the mathematics of the calculation.

Diluting 1 : 5 could alter the cations extracted if soluble salts such as gypsum are present. However, such soils will have low SAR values whichever method is used to prepare the extract. It was not possible to explain why the differences in soil extract preparation should have resulted in different conclusions regarding the relationship of SAR and ESP. The remaining possibility is that the soils in this study differed from those used by Gupta and Abrol (1990) and Rengasamy and Churchman (1999). Ultimately, ease of preparation will favour the use of 1 : 5 extracts, and the contention of this study that SAR (1:5) is useful remains despite this unresolved point.

Amelioration of sodic subsoils can be achieved by the addition of acidifying agents which release calcium into the soil solution, or by the direct addition of calcium as gypsum to the soil (Gupta and Abrol 1990; Prasad and Goswami 1992). The amounts required to cause a significant change in ESP at depth are large and uneconomic. If, as is proposed here, ESP and SAR are not necessarily related, much smaller applications of gypsum may cause a change in the SAR of the soil solution, leading to productivity improvements.

The Natural Gypsum Miners Association in Victoria (pers. comm.) report annual production of 600 000-800 000 t gypsum each year, which strongly suggests Victorian farmers (cropping some 3 Mha dryland) have empirically determined that its addition is beneficial to crop growth. In many cases it is applied to soils with no surface structure problems, and is usually applied at rates of 1-2.5 t/ha, which are insufficient to alter ESP significantly at depth. The evolution of this farming practice strongly supports an alternative explanation such as that advanced here.

A further question arises from the proposition that carbonate formation in these soils is incomplete due to a lack of carbonate. Biomass production has been much increased by modern farming practices and fertiliser application, and it is possible that this has increased root and microbial respiration in the soil (increasing carbon dioxide evolution) and may be causing subsoils to become increasingly hostile to root growth by precipitating calcium and magnesium and causing SAR to rise.

Acknowledgements

The West Wimmera Shire gave permission for the results of its study into sustainable irrigation practices to be used in preparing this paper.

References

Armstrong R, Letts E, Dreccer F, Noy D (2003) The role of 'boron tolerance' in improving adaptation of cereals to highly alkaline subsoils. In 'Solutions for a better environment'. Proceedings of the 11th Australian Agronomy Conference, Geelong, Victoria. (Australian Society of Agronomy: Melbourne, Vic.)

Baxter NM, Williamson JR, Brown AJ (1998) 'Agricultural and horticultural land suitability for West Wimmera Shire Vol I.' (Department of Natural Resources and Environment, Vic.: Melbourne, Vic.)

Gupta RK, Abrol IP (1990) Salt- affected soils: Their reclamation and management for crop production. Advances in Soil Science 11, 224-288.

Lauchli A, Epstein E (1990) Plant responses to saline and sodic conditions. In 'Agricultural salinity assessment and management'. (Ed. KE Tanji) p. 117. (ASCE: New York)

Maas EV, Hoffman GJ (1977) Crop salt tolerance--current assessment. Journal of Irrigation and Drainage 103, 115-134.

Maas EV (1990) Crop salt tolerance. In 'Agricultural salinity assessment and management'. (Ed. KE Tanji) p. 262. (ASCE: New York)

Nuttall J, Armstrong R, Connor D (2003) The effects of salinity, sodicity and soluble boron on wheat yields in the Victorian southern Mallee. In 'Solutions for a better environment'. Proceedings of the 11th Australian Agronomy Conference, Geelong, Victoria. (Australian Society of Agronomy: Melbourne, Vic.)

O'Leary G, Ormesher D, Wells M (2003) Detecting subsoil constraints on farms in the Murray Mallee. In 'Solutions for a better environment'. Proceedings of the 11th Australian Agronomy Conference, Geelong, Victoria. (Australian Society of Agronomy: Melbourne, Vic.)

Prasad R, Goswami NN (1992) Soil fertility restoration and management for sustainable agriculture in south Asia. In 'Soil restoration'. (Eds R Lal, BA Stewart) Advances in Soil Science 17, 37-78.

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

Rengasamy P, Churchman GJ (1999) Cation exchange capacity, exchangeable cations, and sodicity. In 'Soil Analysis: an interpretation manual'. (Eds KI Peverill, LA Sparrow, DJ Reuter) (CSIRO: Melbourne, Vic.)

Shaw RJ (1999) Soil salinity--electrical conductivity and chloride. In 'Soil Analysis: an interpretation manual'. (Eds KI Peverill, LA Sparrow, DJ Reuter) (CSIRO: Melbourne, Vic.)

Snowball K, Robson AD (1983) 'Symptoms of nutrient deficiencies, subterraneum clover and wheat.' (Dept of Soil Science and Plant Nutrition, University of Western Australia: WA)

Victorian Irrigation Research and Advisory Services Committee (1980) Quality aspects of farm water supplies. A report of the Victorian Soil Conservation Authority.

Manuscript received 4 June 2003, accepted 6 July 2004

W. K. Gardner

RMB 7399, Horsham, Vic. 3401, Australia.
Table 1. An example of the calculations used to derive
a weighted average salinity value for the area outside
pivot 1 (Shaw 1999)

EC, electrical conductivity (dS/m)

Depth                           EC       Conv.
(cm)             Texture      (1 : 5)    factor

 0-20           Sandy loam     0.10        11
20-40              Clay        0.12         7
40-60              Clay        0.24         7
60-80              Clay        0.64         7
80-100             Clay        0.80         7
Weighted av.

Depth                         Weight    Weighted
(cm)           E[C.sub.sat]   factor       EC

 0-20              1.10        0.41      0.451
20-40              0.84        0.21      0.176
40-60              1.68        0.17      0.286
60-80              4.48        0.15      0.672
80-100             5.60        0.06      0.336
Weighted av.                             1.92

Table 2. Chemical analyses of irrigation waters used
in the 9 centre pivots studied in the West Wimmera

Pivot no.:                      P1      P2      P3

pH                               7.8     7.7     7.4
EC (dS/m)                        1.4     2.4     1.8
Total salts (mg/L)               840    1400    1100
Bicarbonate (mg/L)               300     370     330
Chloride (mg/L)                  280     540     400
Calcium (mg/L)                    68      75      76
Magnesium (mg/L)                  29      56      45
Hardness (CaC[0.sub.3] mg/L)     290     420     370
Sodium (mg/L)                    150     280     190
Potassium (mg/L)                 5.5     7.7     5.7
Iron (mg/L)                     0.66    0.12    0.32
Manganese (mg/L)                0.02   <0.01   <0.01
Boron (mg/L)                    0.17    0.19    0.13
Bicarbonate (mm)                4.92    6.07    5.41
Chloride (mm)                   7.89   15.21   11.27
Calcium (mm)                    1.70    1.87    1.90
Magnesium (mm)                  1.19    2.30    1.85
Sodium (mm)                     6.52   12.17    8.26
Potassium (mm)                  0.14    0.20    0.15
Bicarb-Ca-Mg (mm)               2.03    1.89    1.66
SAR                              3.8       6     4.3

Pivot no.:                      P5      P6      P7

pH                               7.4     7.4     7.1
EC (dS/m)                          2     2.7     1.8
Total salts (mg/L)              1200    1600     110
Bicarbonate (mg/L)               360     390     510
Chloride (mg/L)                  420     640     310
Calcium (mg/L)                    70      98     100
Magnesium (mg/L)                  51      46      32
Hardness (CaC[0.sub.3] mg/L)     390     430     390
Sodium (mg/L)                    210     340     150
Potassium (mg/L)                 6.1     5.1       4
Iron (mg/L)                     0.21   <0.05   <0.05
Manganese (mg/L)               <0.01   <0.01   <0.01
Boron (mg/L)                    0.17    0.37    0.09
Bicarbonate (mm)                5.90    6.39    8.36
Chloride (mm)                  11.83   18.03    8.73
Calcium (mm)                    1.75    2.44    2.49
Magnesium (mm)                  2.10    1.89    1.32
Sodium (mm)                     9.13   14.78    6.52
Potassium (mm)                  0.16    0.13    0.10
Bicarb-Ca-Mg (mm)               2.06    2.06    4.55
SAR                              4.7       7     3.4

Pivot no.:                      P8      P9      P10

pH                               7.4     7.4     7.7
EC (dS/m)                        2.4     2.4     1.4
Total salts (mg/L)              1400    1400     840
Bicarbonate (mg/L)               400     370     300
Chloride (mg/L)                  530     550     290
Calcium (mg/L)                    57      71      64
Magnesium (mg/L)                  57      59      31
Hardness (CaC[0.sub.3] mg/L)     380     420     290
Sodium (mg/L)                    300     290     180
Potassium (mg/L)                   8     8.1     5.6
Iron (mg/L)                     0.08    0.82     1.3
Manganese (mg/L)               <0.01   <0.01    0.02
Boron (mg/L)                    0.64    0.25    0.14
Bicarbonate (mm)                6.56    6.07    4.92
Chloride (mm)                  14.93   15.49    8.17
Calcium (mm)                    1.42    1.77    1.60
Magnesium (mm)                  2.35    2.43    1.28
Sodium (mm)                    13.04   12.61    7.83
Potassium (mm)                  0.20    0.21    0.14
Bicarb-Ca-Mg (mm)               2.79    1.87    2.05
SAR                              6.7     6.1       4

Table 3. Weighted electrical conductivities at 4 locations of
pivot 4 both inside and outside the irrigated area, and the
derived potential yield loss of white clover this would cause

             Weighted EC (1 m)     Potential yield loss (%)

             Inside    Outside   Inside   Outside   Difference

Location 1    2.98      1.84       19        8          11
Location 2    1.82      1.04        8        0           8
Location 3    2.40      2.10       13       11           2
Location 4    5.00      2.35       38       13          25

Table 4. Weighted electrical conductivities at 2 locations
of pivot 3 both inside and outside the irrigated area, and
the derived potential yield loss of white clover this would
cause

               Weighted EC        Potential yield lost (%)

             Inside   Outside   Inside   Outside   Difference

Location 1    3.06     1.91       20        9          11
Location 2    6.65     3.7        54       26          29
Gale Copyright: Copyright 2004 Gale, Cengage Learning. All rights reserved.