Highlight: A detailed examination was made into the soil the arid and semi-arid regions of Australia (Leigh and Noble, salinity distribution of two Atriplex communities, A. vesicaria and A. nummularia, established at regular spacings on two soil 1969) and other parts of the arid world. They are tolerant to types. The results based on total soluble salts and chloridesdrought and salinity and have high salt content in their leaves suggested that both saltbush species induced significantly (Wood, 1925;Beadle et al., 1957). Our own studies (Sharma et higher salinity in the 0-15 cm soil horizon beneath the bush al. unpublished) on oldman saltbush (A. nummularia) and canopies compared to between the bushes, although A. num-bladder saltbush (A. vesicaria) have indicated the extremely mularia induced significantly higher salinity than A. vesicaria high salt concentration in their leaves and its fluctuation with on both the soils. The pH was also significantly increased under the season. The leaves of these and similar plants, after falling the bushes, but only for the 0-7.5 cm layer. A mechanism for plant-induced soil salinity is proposed by which distinct zones of salt depletion, accumulation and compensation are established. It is suggested that the accumulation of significantly large quantities of salt in the surface layer under the bushes occurred as a result of decomposition of large quantities of salt-rich leaves and fruits. This salt is mainly derived from the soil profile under the plant. Implications of the spatial variability in soil salinity are discussed.

The surface layers of a red earth near Cobar, N.S.W., have been classified into four condition classes by using attributes that are readily observable in the field. These classes provide a reliable estimate of potential dry matter production. Soil samples from the three depths 0-1, 1-5 and 5-10 cm of each class were analysed for various physical and chemical properties. Soil surface resistance was measured in the field. Field measurements of infiltration were carried out at supply potentials of +10 and -40 mm by using the disc permeameter. Differences in physical and chemical properties of the classes occurred mainly in the 0-1 cm layer. Below 1 cm there was very little difference in properties between the four classes. The 0-1 cm depth of the class 1 soils (most productive) contained significantly (P=0.05) higher amounts of organic matter, available nitrogen and exchangeable magnesium than the class 4 soils (least productive). Class 1 soils had higher aggregate stabilities in the 0-1 cm layer and lower soil surface resistance than class 4 soils, indicating more favourable soil physical conditions for seed germination and emergence. Aggregates of all classes underwent considerable dispersion in water when subjected to mechanical stress. At a supply potential of +10 mm, class 1 soils had a significantly (P=0.05) higher infiltration rate than class 4 soils. However, at a supply potential of -40 mm, there was no difference in infiltration rate between classes. Scanning electron microscope (SEM) micrographs showed that cryptogamic material on the surface of class 1 soils resulted in a porous, well aggregated structure compared with the dense, compacted structure of class 4 soils. The differences in physical and chemical properties support earlier micro-morphological evidence that the classes 1-4 essentially represent an erosional sequence from no erosion (class 1) to strongly eroded (class 4). The importance of these physical and chemical properties of different soil classes in determining productivity of rangeland soils is briefly discussed.

Criteria for the classification of Red Earth soils into groups reflecting differences in their productive potenrial were investigated on three sites near Cobar, New South Wales. Five features were shown to be useful by cluster and discriminate analysis of an extensive data set. A field-based classification using six very similar criteria was also tested and shown to agree with the cluster method. Laboratory tests for stability and fertility of the surface soil layer from each class indicated a significant difference between classes, and a narrow range of values for an individual class. Herbage production on quadrats over the range of soil surface condition also consistently reflected the classification. When the method was tested with three personnel previously unfamiliar with the criteria, good agreement with the original experimental set was obtained.

This article outlines the exploration and settlement of poplar box lands which followed the European occupation of the Australian continent in the mid-19th century, draws attention to the various fragments of historical information on substantial changes in the soils, vegetation and biota, and describes the role of European livestock (sheep, cattle, goats, rabbits) and fire in these changes.

A rational analytical approach to the restoration of degraded rangelands is proposed in which degradation is described in terms of loss or diminution of plant habitat favourability. The approach recognizes that fertile patches or zones are maintained by the action and interaction of abiotic and biotic transfer processes. The identity and character of these processes can be deduced from data derived from integrated vegetation and landscape surveys. Ecosystem function and disfunction can be identified and geographically referenced by this procedure. Finally, restoration practices are aimed at reconstituting the landscape processes 'normal' to the ecosystem in question.