Management of dryland salinity in Australia will require changes in the design and utilisation of plant systems in agriculture. These changes will provide new opportunities for livestock agriculture. In areas already affected by salt, a range of plants can be grown from high feeding value legumes with moderate salt tolerance through to highly salt tolerant shrubs. A hectare of these plants may support between 500 and 2000 sheep grazing days per year. The type of plants that can be grown and the subsequent animal production potential depend on a range of factors that contribute to the ‘salinity stress index’ of a site, including soil and groundwater salinity, the extent and duration of waterlogging and inundation, the pattern and quantity of annual rainfall, soil texture and chemistry, site topography and other site parameters. Where the salinity stress index is high, plant options will usually include a halophytic shrub that accumulates salt. High salt intakes by grazing ruminants depress feed intake and production. Where high and low salt feeds are available together, ruminants will endeavour to select a diet that optimises the overall feeding value of the ingested diet. In areas that are not yet salt affected but contribute to groundwater recharge, perennial pasture species offer an opportunity for improved water and salt management both on-farm and at the catchments. If perennial pasture systems are to be adopted on a broad scale, they will need to be more profitable than current annual systems. In the high rainfall zones in Victoria and Western Australia, integrated bioeconomic and hydrological modelling indicates that selection of perennial pasture plants to match requirements of a highly productive livestock system significantly improves farm profit and reduces groundwater recharge. In the low to medium rainfall zones, fewer perennial plant options are available. However, studies aiming to use a palette of plant species that collectively provide resilience to the environment while maintaining profitable livestock production may also lead to new options for livestock in the traditional cropping zone.

Existing perennial plant-based farming systems are examined within 4 climatic zones in southern Australia (western winter rainfall, south-eastern low to medium rainfall, south-eastern high rainfall and northern summer rainfall) to assess their potential to improve the management of dryland salinity. If profit is to be the primary driver of adoption, it appears that the available options (lucerne and other perennial pastures, farm forestry, saltland pastures and forage shrubs) will fall short of existing hydrological targets with the exception of the higher rainfall zones. In the 3 eastern zones, the need to preserve fresh water flows to permanent river systems places limitations on the use of perennial plants, while the higher proportion of regional groundwater flow systems increases response times and heightens the need for regional coordination of effort. In the western zone, the prevalence of local and intermediate ground water flow systems increases effectiveness of individual action. Research into new perennial land use systems has been characterised by an emphasis on water use over profit resulting from poor dialogue between paddock, farm and catchment scales. Exploring the water use implications of land use systems that are potentially viable at farm scale is a more promising approach than focusing on the opportunity cost of catchment scale intervention. Perennial plant-based farming systems present both threats and opportunities to native biodiversity. The major threat is the introduction of new environmental weeds. The opportunities are potential improvements in vegetative cover, food sources and habitat for the native biota, but only where nature conservation goals can influence the structural complexity, composition and location of new land use systems.

Half-sib families in the AT98 Phalaris aquatica × P. arundinacea × P. aquatica backcross population bred for acid soil tolerance were compared for establishment, persistence, and yield with phalaris (P. aquatica L.) controls and cocksfoot (Dactylis glomerata L. cv. Porto) at 4 sites in south-eastern Australia with the aim of selecting the parents of a new cultivar. The sites had strongly acid soils but differed in parent material, pH profile, soil fertility, and suitability for phalaris. Establishment by AT98 was clearly superior to all phalaris controls and similar to cocksfoot in an acid soil high in Al to depth at Chiltern, north-eastern Victoria, after sowing in early spring 2000. It was considered likely that better establishment by AT98 was due to its higher Al tolerance. In contrast, little variation in establishment was observed at 3 other sites sown in late autumn 1999, possibly due to a longer period free of moisture stress compared with the later sown Chiltern site. Once established, the control cultivars of phalaris at the autumn-sown sites in general persisted and yielded similarly to the mean of the AT98 families. Significant family variation was observed and predicted heritability on a family mean basis was high for persistence measured as basal frequency and moderately high for yield in the third year. Family by site interaction was relatively low for both attributes. A cultivar based on the best families should give more reliable establishment on acid soils high in Al under conditions where rapid root growth to depth is needed for survival, and give more flexibility of sowing date on these soils. Its best performance relative to cv. Landmaster in terms of third-year yield was predicted to occur on granite-derived soils in north-eastern Victoria.

Persistence of an aluminium (Al)-tolerant phalaris F1 hybrid (Siro 1146 Phalaris aquatica × P.�arundinacea) was compared with that of 3 more Al-sensitive phalaris (P. aquatica L.) lines (cv. Australian and the progenitors of cvv. Sirosa and Sirolan) at 2 sites (Strathbogie and Baddaginnie) in north-eastern Victoria with strongly acidic soils 20 years after sowing. Soil pHCa and extractable aluminium (AlCa) were also measured to 1 m depth under Siro 1146 and annual grass pasture at each site. All grass treatments contained volunteer subclover. Siro 1146 persisted better than the other lines (P<0.05) at Strathbogie where the soil contained high AlCa concentrations down to 50 cm depth. Soils should be sampled and tested to at least this depth to determine their suitability for phalaris. At this site Australian phalaris persisted better than the progenitors of Sirosa and Sirolan (P<0.01), probably because Australian has more spreading ability and tolerance of set stocking than the winter active lines. At Baddaginnie, the soil contained lower concentrations of AlCa below 20 cm depth than at Strathbogie and persistence of the 3 Al sensitive phalaris lines was good despite the lower rainfall. The less drought-tolerant Siro�1146 persisted poorly at Baddaginnie, but had high ground cover due to high lateral spread of the survivors. At Strathbogie, the 10–20 cm layer of soil under Siro 1146 had a higher pHCa and lower AlCa than that under the other 4 phalaris and the annual grass treatments. Soil under Siro 1146 also had higher pHCa and lower AlCa down to 40�cm�depth compared with the annual grass treatment, the differences being significant in the 20–30 cm layer. These differences were reversed at 50–100 cm, but only the effect on Al was significant. Although soils were not sampled at the beginning of the experiment, adequate replication and randomisation of the grass treatments showed that it was statistically improbable for observed final differences to be due to initial soil differences. The final differences may be due to greater amounts of nitrate being taken up from the upper layers by Siro 1146 over its long growing season compared with the annual grasses, leading to greater nitrate leaching from the upper layers and greater nitrate uptake from lower layers under the annuals (i.e. spatial separation of acid generation and consumption processes within the profile). Another possible reason for the greater acidification below 50 cm by the perennial is that its roots may have taken up more cations in this zone. Because of rising concern in some sectors of the public about the off-site, environmental effects of agriculture, the future role of more acid-tolerant phalaris cultivars growing near native vegetation in the high rainfall zone is discussed.

Grazing management strategies that included resting or intensely utilising pasture on a seasonal basis were compared for their effects on phalaris production and plant frequency (persistence). Experiments were established at 4 on-farm sites (Cootamundra ‘old’, Cootamundra ‘new’, Springhurst and Cavendish) in southern New South Wales and Victoria that had previously been sown to phalaris and were grazed by sheep. At each site, 8 core treatments and extra locally determined treatments were initially imposed in 1993–94 on 2 spatial replicates. In order to determine and describe any year of start effects, treatments were applied again in 1994–95 to plots that had been maintained as controls. The phalaris component of the pastures varied from a minor component to the dominant component depending on site (11–54%). Measurements of botanical composition, available herbage and plant frequency were made between September 1993 and September 1996. Of the core treatments, autumn closure (Cootamundra old and new), winter closure and grazing between defined levels of available herbage (mob stocking) during autumn–winter (Cootamundra old and new, Springhurst), were the most effective in either maintaining or increasing phalaris herbage mass compared to the continually grazed control treatment. In addition, the frequency of phalaris was higher than the control at each of these sites for the autumn–winter mob stocking treatment. These treatments had no effect at the Cavendish site where phalaris was a minor component of the pasture. Rotational grazing, imposed at 2 of the sites (Cavendish and Cootamundra new), led to an increase in phalaris herbage mass compared to continual grazing. A further treatment aimed at encouraging phalaris seedling recruitment by using an extended spring rest until seed fall in summer followed by a rest after the autumn break was imposed at the Cootamundra old site. This treatment increased phalaris herbage mass but did not result in seedling recruitment. The results emphasise the need for periods of rest when buds are regenerating and tillers developing over the autumn–winter period for phalaris pastures in summer-dry environments.