Measurements of naturally occurring stable isotopes of hydrogen (2H) and oxygen (18O) were used to determine the importance of creek waters to river red gums (Eucalyptus camaldulensis Dehnh., Myrtaceae) on the banks of creeks in the semi-arid and saline Chowilla floodplain region of South Australia. The 2H and 18O concentrations of water sampled from trees were significantly different from those of the creek water, indicating that, generally, trees were not obtaining all their water from the creek. This was found even in trees overlying highly saline groundwater. Thus, saline groundwater and/or soil water were important water sources for the trees studied. The results indicated that the water relations of these trees are complex, and the trees may be less affected by changes in creek flow and/or salinity than was previously anticipated.

Particle-size analysis (PSA) methods to be used in routine situations need to be rapid, require no prior information about the sample and give precise results over a wide range of soil textures. Effects of three physical dispersion and two fine-fraction determination methods on PSA results were investigated over a wide range of soil textures to find the most appropriate technique for routine PSA. Interactions between physical dispersion and fine-fraction determination methods were also investigated. The reciprocating shaker produced significantly lower silt and fine sand and higher coarse sand (and clay, although not significant) values than either drink mixer or end-over-end shaker dispersions. This result was interpreted as the reciprocating shaker giving the most effective dispersion, with aggregated clay being dispersed to primary particles while coarse sand was not fragmented to fine sand or silt. The end-over-end shaker did not reliably disperse a heavy clay soil, and so cannot be recommended for routine use where similar soils may be encountered. When considered over all soils and dispersion methods, hydrometer clay and clay + silt values were higher than pipette values. These results were due to the effective depth of the hydrometer being greater, on average, than the depth of the pipette. However, there were interactions between dispersion and fine-fraction determination methods for the clay and clay +silt classes. Hydrometer values were greater than pipette values with drink mixer and end-over-end shaker dispersion, but were similar with reciprocating shaker dispersion. For the clay fraction, inferior dispersion given by the drink mixer and end-over-end shaker resulted in a significant mass of particles between the sampling depths of the pipette and hydrometer, causing the higher hydrometer values. For the clay +silt fraction, both drink mixer and end-over-end dispersion methods fragmented sands to a size which was recorded by the hydrometer but not the pipette. These interactions highlighted the requirement for effective dispersion where clay and silt are determined by the hydrometer, and may explain some of the conflicting observations of the precision of the hydrometer compared with the pipette. Reciprocating shaker physical dispersion combined with the hydrometer fine-fraction determination method was found the most appropriate PSA method combination for use in a routine situation.

Infiltration and deep drainage fluxes are difficult to measure directly in slowly permeable soils under furrow-irrigated upland crops such as maize (Zea mays L). This paper combines a solute mass balance model and a furrow irrigation advance model to provide an estimate of these fluxes, indirectly from simpler measurements. The models were applied to a newly cleared sodic duplex soil from the Burdekin River Irrigation Area, north Queensland, where no field measures of infiltration and deep drainage were available. The study site was sown to consecutive furrow-irrigated crops after clearing. In applying these models, measures of soil and irrigation water chloride, irrigation water applied, furrow geometry and irrigation advance were required. Estimated infiltration and deep drainage decreased with distance down the furrow from 1044 and 98 mm year-1 at 50 m to 966 and 0 mm year-1 at 260 m. In an area that received an application of gypsum (20 t ha-1) prior to planting the second crop, values ranged from 1617 and 200 mm year-1 at 50 m to 1370 and 70 mm year-1 at 260 m. Infiltration did not satisfy the estimated soil water deficit unless gypsum was applied. Where gypsum was applied, infiltration exceeded the soil water deficit and deep drainage increased. Evapo-transpiration rate and maize yield also increased. If correct, these results have local and regional implications for irrigation design and management.

No abstract

Four experiments were conducted by using individual aggregates and ground samples of different soils and a range of wetting pre-treatments to investigate the effect of wetting on the desorption moisture characteristic of Vertisols. Wetting rate was varied by manipulating the energy of the water used to wet up the soil. Fast wetting was found to increase moisture uptake in Vertisols over a wide range of matric potentials (to -38 MPa in one soil). At high matric potentials, additional water uptake ranged up to 0.26 g g-l. These results were attributed to the creation of planar voids within soils between small water-stable aggregates on fast wetting. These voids retained water at matric potentials as low as -90 kPa. At this potential, 40% of additional water taken up on fast wetting was still retained in one soil. At -1500 kPa, fast wet soils retained up to 1.2 times more moisture relative to slow wet soils. The effect of wetting at potentials of -1500 kPa and less appeared to be related to the smectite content of the soil clay fraction and the degree of order of clay quasi-crystals. The soil which showed the greatest effect of fast wetting at these potentials had a less well ordered clay micro-structure. These findings have relevance to field studies, as variations in maximum field soil moisture content, with depth, could be reproduced by tension wetting pre-treatments. This result indicates that reduced rate of subsoil wetting may be responsible for the decrease in maximum field water contents with increasing soil depth, which has been observed in Vertisols. Such profiles have previously been attributed to the effect of overburden and stress potentials alone.