Our Valuable Native Grasslands, Better Pastures Naturally
Proceedings of the Second National Conference of the Native Grasses Association

Approaches to managing the water balance of grasslands in southeastern Australia

WH (Bill) Johnston

DLWC Centre for Natural Resources
PO Box 189 Queanbeyan NSW 2620

Abstract

Managing pastures to manage the water balance is an increasingly important goal in southern Australia. The composition and management of grasslands affect three interdependent terms of the water balance: water use, runoff and lateral flow, and deep drainage.

Water use is linked to the evaporation potential of the atmosphere, which is high in summer and low in winter. Water use in summer creates a deficit that delays when field capacity is reached in autumn and winter and reduces the potential for deep drainage. In undisturbed grasslands, runoff is an important water balance component that results in positive off-site impacts. It also reduces the potential for deep drainage. Unless there is a compensating increase in water use, cultivation results in increased infiltration, soil wetness, and deep drainage at the expense of runoff.

Maintaining native grasslands on catchment uplands can result in production and catchment protection benefits. Set stocking at low stocking rates is particularly detrimental and should be avoided. Minimizing the effects of selective grazing, allowing pastures to recover after grazing, and managing the flush of growth that occurs in pastures in spring requires a flexible and strategic approach to grazing management.

Introduction

In southern Australia, grasslands of one form or another dominate the landscape. The spectrum of grasslands includes at one extreme, highly modified and disturbed, simple communities of annual plants (including crops), and at the other, relatively undisturbed, diverse communities that contain a range of perennial grasses, shrubs and trees. Across this spectrum, the proportions of perennial plants, and the seasons when they grow most actively, are key characteristics affecting their water use patterns.

Water use is a passive, climate-driven process. On the one hand, it requires a water source (within the soil) linked by a conductive network of roots or soil pores, to evaporative surfaces (green leaves or the soil surface itself), whilst on the other, an atmosphere capable of drying the leaf or soil surface. Interplay between supply and demand, moderated by the process of water movement through the system, including resistances in soil and plants, determines when, how quickly and for how long the vegetation continues to grow and how much water it uses.

This paper provides an overview of the various terms in the water balance, and outlines the role of productive, grazed native grasslands in a catchment protection/sustainability context.

The water balance and its components

The water balance accounts for the fate of rain falling on the landscape. It is expressed as an equation such as:

Rainfall = evaporation + interception + change in soil water content + runoff and lateral flow + deep drainage

Water use and evaporation

Evaporation is the largest term in the water balance. In the 400-600 mm rainfall zone of southern Australia, evaporation from a water surface exceeds rainfall by a factor of about 2 on an annual basis, and is the dominant pathway for water loss. Although the climate is summer-dry, with about half of the total evaporation occurring between December to February, a surplus of rainfall over evaporation is common in winter and early spring. Two interacting processes determine rates of water use:

In summer, it depends on the amount of green leaf present, the depth and density of root systems, and the various adaptive responses of plants to high temperature, solar radiation and internal water stress, such as leaf rolling, waxing and leaf senescence.

In winter, when soil water is in ample supply, resistances to water loss in actively growing plants are low, but rates of water use are limited by low evaporation potential. For grasslands containing a low proportion of winter growing species, dormancy and a build-up of mulch from summer reduce water loss further. Low water use in winter prolongs water availability in early summer, which is important for summer-growing plants.

Interception

Interception is the amount of rainfall that never reaches the ground - it evaporates directly from plant canopies. It is an often-forgotten component of the water balance, but depending on foliage density, it accounts for 20-30% of the rainfall in forests and dense woodlands, and some 1-10% in grassland and pasture (O'Loughlin 1988). Direct evaporation from tree canopies is higher than for pastures because of their exposure to turbulence.

The biggest impact of tree clearing is to increase the amount of rainfall reaching the soil surface; and the biggest impact of reafforestation is to increase interception loss (Holmes and Sinclair 1986). Interception by needle-type foliage is 8-10% higher than native forest (Dunin and Mackay 1982), and large-scale planting of Pinus spp. has been noted to considerably reduce water yield (O'Loughlin 1988).

Changes in soil water content

Changes in the amount of water held by the soil account for a proportion of the rainfall only if the soil is capable of wetting up from a drier state. Soils at `maximum wetness' act only as conductors - water moves through them vertically or laterally under the influence of gravity, at rates determined by their porosity.

Soils reach the limit of their dryness in late summer, and their maximum water holding capacity (or field capacity) in late winter. These two points define the available water holding capacity. Deep friable soils may have an available water capacity of greater than 150 mm to 1 m depth. Shallow, stony soils may only hold 50 mm. Soils that hold less water reach field capacity more rapidly in winter, and dry out more quickly in summer, thus contributing a greater proportion of their water balance to either runoff or deep drainage.

Changes in topsoil depth due to past soil erosion, and increased soil density resulting from loss of organic matter, and grazing or cultivation, reduce the amount of water that a soil may hold. For soils with a low infiltration capacity, this would be reflected in increased runoff or waterlogging. For highly permeable soils it would increase deep drainage rates.

Land managers have no control over when rain falls or how much is received. However they can influence how much of the rainfall the soil can hold before field capacity is exceeded in autumn and winter, and runoff or deep drainage occurs. Plants that use more water in summer, or grazing strategies that encourage greater water use, create a larger pre-winter soil water deficit, which delays attainment of field capacity (Johnston 1993).

In southern Australia, soils usually wet up in autumn in a step-wise fashion in response to relatively heavy (>25 mm in up to 4 days) rainfall events. Figure 1 shows that at Wagga Wagga NSW, the likelihood of rainfall exceeding 25 mm/day during March, April or May is about 25%. For these months combined, it exceeds 60%. Between 1^st March and 1^st July, the chance that rainfall will exceed 25 mm over a 4-day period is more than 90%.

Soils that develop only a small water deficit in summer, for example those that have been fallowed, or whose surface was covered by mulch from the previous spring quickly reach field capacity. Under these circumstances the risk of deep drainage and leaching of soluble plant nutrients (particularly N) is high.

The combination of a low pre-winter deficit, and a high likelihood of heavy rainfall in autumn, are major drivers of soil acidification, waterlogging, deep drainage and salinity - processes that threaten the sustainability of agriculture throughout temperate Australia.

Figure 1. The chance of rainfall exceeding 25 mm over up to 7 days at Wagga Wagga for March (_), April (_), May (_), and March to May (_) and March to June inclusive (_).

Runoff and lateral flow

Runoff and lateral flow are normal processes that redistribute water, sediment and nutrients over the landscape. This often results in patterning of the vegetation (Johns 1981, Birch et al. 1987). Runoff occurs when either the soil's capacity to hold water is exceeded because it is wet (saturation-induced runoff) - usually in winter - or when rainfall intensity exceeds the soil surface infiltration capacity (infiltration excess runoff) - usually in the warmer months.

Lateral flow is another form of runoff that occurs when an impermeable layer impedes downward water movement. This is common for sloping duplex soils with well-drained surface horizons overlying relatively impermeable subsoil. If both lateral and vertical water movement is impeded, water ponds within the soil, resulting in the development of perched water tables.

Surface disturbance, such as contour cultivation and deep ripping has a major impact on runoff, because depressions store water, allowing more time for it to infiltrate. In southern NSW, runoff reductions of 20-80% have been noted for disturbed, sown catchments compared to undisturbed native grasslands (Adamson 1976; Costin 1980). If runoff reductions are not matched by an equivalent increase in water use, soils become wetter, increasing lateral flow, waterlogging or deep drainage.

Deep drainage

Deep drainage occurs when water moves down the soil profile beyond the reach of plant roots. Deep drainage rates are higher for shallow, stony, porous soils than for deep clayey soils. Deep drainage may occur in less porous soils if they stay wet for long periods, evaporation rates are low, or there is lateral water movement from up-slope.

Water draining through the landscape mobilises nutrients and salt, which finds its way to local and regional groundwater aquifers. In southeastern Australia, water tables are rising at a rate of 4-10% of the annual rainfall (Johnston et al. 1999). This is a major sustainability issue in the Murray-Darling Basin where more than 9 million hectares are at risk from increasing salinisation due to- high water tables and saline discharges (PMSEIC 1999).

Interdependency of water balance components

Components of the water balance are interdependent. For instance, if runoff is less, the water balance predicts a compensatory proportional change in some other component. Figure 2 depicts this interdependence using data for the Axe Creek catchment (about 300 km²) near Bendigo Victoria (after MDBC 1987).

(In calculating the impact of sown pasture, it was assumed that runoff would decline to about 1% of post-clearing runoff (Adamson 1976; Costin 1980) while water use by the sown summer-dormant pasture would remain about the same. This would be the case for pastures that were set stocked over summer (Clifton et al. 1996)).

Figure 2. Gross effects of land clearing on the water balance of the Axe Creek Catchment, near Bendigo, Victoria (left). The graph on the right provides a more detailed view of changes in the interception, deep drainage and runoff components.

Before it was cleared, interception and runoff were significant components of the water balance of Axe Creek. Clearing reduced interception loss, and evaporation. This resulted in a large increase in runoff and a small but significant increase in deep drainage. If the catchment were sown with a pasture that did not use significantly more water, increased detention storage resulting from cultivation, would result in less runoff, and increased deep drainage.

Managing to manage the water balance

Although the water balance may seem to be complex, in a land management sense it resolves itself into two main issues:

Soil dryness in late summer determines its capacity to take up water in autumn and winter.

Once this capacity is exceeded, excess rainfall must appear as either runoff (including lateral flow), deep drainage or waterlogging (excessive saturation). The impact of waterlogging and the off-site impact of deep drainage are usually negative.

The impact of clearing and agricultural development on components of the water balance

In their pre-European state, the forests and open woodlands of southern Australia were dominated by perennial plants with a capacity for year-long transpiration (Specht 1972). C₄ species, common in dry, light-saturated habitats, owed their success to a variety of water-saving strategies, coupled with efficient water use and tolerance of water stress. C₄ grasses and other C₄ woody and herbaceous species were more abundant in hotter, drier environments west of the Great Dividing Range (Johnston 1996). Perennial, winter-growing C₃ species prefer shaded or moist habitats and achieve high growth rates by spending water rapidly (i.e. they are relatively inefficient). These occurred in greatest numbers in high rainfall areas along the Great Dividing Range, and elsewhere within forest understorey (Moore 1953; Hattersley 1983). C₃ grasses depend on root penetration and moisture acquisition to supply their water requirements and rely on dormancy to escape periods when their moisture supply is exhausted.

Fertility-responsive, annual C₃ grasses and herbs enjoy a strategic advantage as grassland invaders and they have become widely dispersed. Annuals germinate in response to rain in autumn - a time when growth rates of summer-growing species decline. For pastures grazed over summer, there is usually a high proportion of bare ground at this time, which allows them to germinate with little competition.

Annuals reach peak biomass in spring. Their ground cover denies recruitment opportunities to species that set seed the previous summer and which need increasing soil temperature for germination and establishment. The annuals set copious quantities of seed and senesce as evaporation rates increase, so they rapidly deplete soil water reserves. This further disadvantages perennial grasses that depend on stored soil water for growth. Only summer active perennial species that are well adapted to aridity, or summer-dormant perennials that do not need access to large amounts of stored soil water persist well in this modified environment

Competition for both space and light in spring, consistently reduced soil water availability in early summer. High grazing pressure in summer and autumn and lack of recruitment opportunities leads to a seemingly irreversible switch in botanical composition in favour of annual species, and summer dormant C₃ perennials (Moore 1973; Whalley et al. 1978). Species that escape being grazed because they are unpalatable or they are small and inconspicuous, and species capable of being moved around by sheep, are favoured by set-stocking.

These include low-growing species of wallaby grasses (Austrodanthonia), prostrate-growing weeping grass (Microlaena stipoides), bent grasses (Agrostis) and Chilean needle grass (Nassella neesiana), and various coarse, unpalatable, tussock-forming species such as species of corkscrew grass (Austrostipa), Poa, and serrated tussock (Nassella trichotoma). A large number of herbaceous species such as Erodium (crowfoot), thistles, heliotrope and capeweed also invade over-grazed patches.

The change from tall grasslands containing summer active species, to short-growing grasslands containing species that were mostly summer dormant, resulted in a switch in the water use pattern. Soils became drier in late spring, and wetter in autumn (Johnston et al. 1999).

As these modified grasslands do not fully utilise their available resources in wet years, the opportunity exists for annual C₄ grasses and broad-leafed species to gain a foothold. Following drought or fire, or if the replacement grassland loses its vigour, or if other changes occur such as fertility rundown or the soil acidifies, then secondary invasions of perennial C₄ species may follow. The species involved are invariably adapted to dry, infertile relatively harsh habitats. Examples include wire grass, rat's tail grass and lovegrass (Aristida, Sporobolus, and Eragrostis); prolific seeders that recruit efficiently such as windmill grass, canary grass and bottlewashers (Chloris, Setaria and Enneapogon); or those that can avoid grazing because of their low stature (eg. couch grass and redgrass [Cynodon and Bothriochloa]) or unpalatability (e.g. lovegrass and Coolatai grass [Eragrostis curvula and Hyparrhenia hirta]). Although many of these species may have weedy characteristics, their invasion, which is facilitated by disturbance, bare ground and incomplete water use, would tend to increase the pre-winter soil water deficit and reduce deep drainage.

Landscape considerations

Landscapes vary in their potential to support stable productive ecosystems. A landscape-based plan that recognises physical and productive limitations and the types of grasslands that naturally occur is an essential management tool.

It is important to separate areas where it is desirable to maintain stable productive grasslands from areas that can safely be more intensively managed. Management can then be targeted more closely to meeting the specific needs of the range of all land management units. It can achieve multiple outcomes in term of productivity, landscape stability and biodiversity, as well as on- and off-site benefits, including water management.

Soils on hill lands are typically shallow, leached and permeable. They partition a high proportion of their water balance to either runoff, lateral flow or deep drainage and their hydrology is very responsive to changes in land use and vegetation cover. Although hill lands may be less productive in an agricultural sense, they are particularly important for their role in maintaining the water yield, quality and other habitat values of catchments (Garden et al. 2000). Deterioration in the condition of hill lands, or changes in their hydrology have river basin-wide implications.

Because they are generally less productive than low-slope lands, hill lands are often seen to not warrant the investments needed in order to manage them well. In the higher rainfall zones (> 500 mm), large areas in the past were fertilised and sown to exotic pasture grasses and legumes. As fertiliser applications became less economic and soil pH declined, and under the relentless pressure of set-stocked grazing, many previously `improved' pastures have reverted to grasslands dominated by unpalatable species, or weedy annuals. Factors contributing to their deterioration have included lack of, or inappropriate subdivision, poor ground cover in summer, tracking by livestock and nutrient transfers to stock camps.

Across large areas of the uplands of southeastern Australia, grasslands have reached the point where desirable perennial grasses have been considerably weakened, or have gone forever. There is a need to develop a range of native grasses for reseeding into areas where they have been lost, and management strategies that arrest and reverse current degradation trends. Because of their importance in a catchment protection context, and the difficulty of reclaiming the diversity that underpins this role, hill lands that are presently in reasonable condition need to be managed with care.

Managing grasslands for multiple outcomes

Grasslands that occur on hilly landscapes need to be managed to preserve and enhance their biodiversity and catchment protection role (e.g. Simpson 2000). Unfortunately, for many grassland types (particularly those containing C₄ grasses), there is scant information about how to achieve productivity gains without compromising their integrity.

Managing the water balance is an increasingly important goal right across the landscape. For pastures to mimic the water balance of pre-European plant communities, they must maintain their capacity to generate runoff, and to transpire water in summer and early autumn. Grasslands containing C₄ perennial grasses respond to summer rain, have a strong capacity to use water over summer, and they create substantial soil water deficits. Water use by C₃ native grasses and sown pasture species, such as phalaris and cocksfoot (Phalaris aquatica and Dactylis glomerata), is limited in summer by low growth rates caused by leaf senescence and dormancy and often their regenerative buds are protected by accumulated mulch. Continuous grazing compounds these issues because any green leaf remaining is selectively grazed.

Maintaining the biodiversity, stability and productivity of grazed grasslands depends on the grazing pressure that is applied, how long it is applied for and the time of the year it is grazed.

For the feed-base to last, set stocking generally requires low grazing rates. This is very damaging to grasslands that contain an array of species having varying palatability and growth patterns. Selective grazing leads to palatable species becoming over-grazed, to the advantage of less palatable species. Livestock may also only graze small areas and create a pattern of overgrazed and under-grazed patches. This leads to bare ground, weed invasions and other undesirable consequences. Patchiness of grazing is a sign that grazing rates are too low. Increasing grazing rates to overcome patch grazing will usually require adoption of a rotational form of grazing, which is far more beneficial for the grassland.

Grazing pastures with high stock numbers for short periods minimises selective grazing and patterned grazing, and promotes green leaf recovery which is important for summer water use. As well as achieving more uniform grazing, high grazing pressure redistributes nutrients more evenly across the pasture. For pastures containing summer-growing species, spring is a critical period. Heavy, tactical grazing in early spring reduces competition from the usual flush of growth by annual species. Resting during early summer encourages flowering and seed-set, particularly in above-average rainfall years, which is important to preserve the seed pool.

It is vital to maintain the nutrient capital of grazed grasslands by applying fertiliser. However, the loss of surface-applied nutrients in runoff is an important catchment management issue. Fertiliser should be applied when the risk of high intensity rainfall is low in autumn, and at rates that result in gradual increases in pasture production. A grazing regime where stocking rates increase in line with pasture production will help maintain a balance between annual and perennial species, while capturing the economic benefits of increasing soil fertility (Simpson and Langford 1996).

Although runoff from pastureland occurs most often in winter when ground cover is high and soils are saturated, high intensity storms in late spring and summer may generate high peak flows that considerably increase the potential for soil erosion. The erosion risk increases markedly as the amount of cover provided to the soil surface by standing vegetation and litter decreases below about 70% (Costin 1980). It is also higher where stock tracks and overgrazed patches allow runoff to concentrate and flow at a high velocity.

Conclusions

The growth and water use characteristics of grasslands have important impacts on catchment health. Managing the water balance is an increasingly important goal right across the landscape. The main characteristics that determine the impact of grasslands on the water balance are:

Water use in summer. Grasslands that use water in summer, dry the soil and reduce the potential for deep drainage to a greater extent than pastures that are summer dormant or senescent.
Impact on runoff. Cultivation increases infiltration at the expense of runoff. If this is not balanced by an equivalent increase in water use, it will result in soils becoming wetter, and lead to higher rates of deep drainage.

Conversion of summer-growing grassland to a summer-dormant sown pasture is likely to reduce runoff, and markedly increase deep drainage.

Land-use planning must be linked to land capability, and a landscape-based plan is an essential management tool. Grazing management should aim to maintain the diversity of species within grasslands, as well as their capacity to use water in summer, and their ability to generate runoff.

This requires a flexible and strategic approach to grazing. High grazing pressure for short periods followed by extended rest periods may be the most appropriate management strategy for diverse grasslands. Spring is a critical period for fertilised pastures because competition from fertility-responsive annuals may reduce persistence of summer active perennial grasses.

Heavy tactical grazing can be used in early spring to reduce competition. Resting in early summer when perennials are flowering and setting seed is important for maintaining their capacity to recruit. Ground cover should be maintained above 70% to minimise the risk of soil erosion and reduce the risk of invasions by undesirable species.

Acknowledgements

Bill Semple's helpful comments on earlier versions of the manuscript were appreciated.

References

Adamson C.M. (1976). Some effects of soil conservation treatment on the hydrology of a small rural catchment at Wagga. Journal of Soil Conservation NSW 32, 230-249.
Burch G.J., Bath R.K., Moore I.D. and O'Loughlin E.M. (1987). Comparative hydrological behaviour of forested and cleared catchments in southeastern Australia. Journal of Hydrology 90, 19-42
Clifton C.A., Taylor J.M., and Schroder P.M. (1996). Pasture management, the water balance and dryland salinity management. In: `Perennial Pastures for Recharge Control: Report on a Workshop'. pp. 2-26. (Occasional Paper 04/96; Land and Water Resources Research and Development Corporation: Canberra.)
Costin A.B. (1980). Runoff and soil and nutrient losses from an improved pasture at Ginninderra, Southern Tablelands, New South Wales. Australian Journal of Agricultural Research 31, 533-546.
Dunin F.X. and Mackay S.M. (1982). Evaporation by eucalypt and coniferous plant communities. In: `The First National Symposium on Forest Hydrology, Melbourne, 11-13 May 1982' pp. 18-25. (The Institution of Engineers, Australia).
Garden D.L., Johnston W.H., Simpson P.C. and Langford C.M. (2000). Pasture management challenges in higher altitude cool season rangelands of NSW. Proceedings of the 11th Australian Rangeland Society Conference - Broken Hill pp. 98 - 102 (Australian Rangeland Society).
Hattersley P.W. (1983). The distribution of C₃ and C₄ grasses in Australia in relation to climate. Oecologia (Berlin) 57, 113-128.
Holmes J.W. and Sinclair J.A. (1986). Streamflow from some afforested catchments in Victoria. In: `Hydrology and Water Resources Symposium, Griffith University Brisbane, 25-27 November 1986' pp. 214-218. (The Institution of Engineers, Australia)
Johns G.G. (1981). Hydrological processes and herbage production in shrub-invaded poplar-box (Eucalyptus populnea) woodlands. Australian Rangelands Journal 3, 45-55.
Johnston W.H. (1996). The place of C₄ grasses in temperate pastures in Australia. New Zealand Journal of Agricultural Research 39, 527-540.
Johnston W.H., Clifton C.A., Cole I.A., Koen T.B., Mitchell M.L., and Waterhouse D.B. (1999). Low input grasses useful in limiting environments (LIGULE). Australian Journal of Agricultural Research 50, 29-53.
Johnston W.H. (1993). Water use issues in temperate agriculture. In: `National Conference on Land Management for Dryland Salinity Control. Bendigo, September 1993'. pp. 32-43. (Centre for Land Protection Research: Bendigo.)
Moore C.W.E. (1953). The vegetation of the south-eastern Riverina New South Wales I. The climax communities. Australian Journal of Botany, 1, 485-547.
Moore R.M. (1973). Australian grasslands. In `Australian Grasslands.' (Ed. R.M. Moore.) pp. 85-100. (Australian National University Press: Canberra.)
Murray-Darling Basin Ministerial Council (1987). `Murray-Darling Basin Environmental Resources Study'. (State Pollution Control Commission: Sydney (for MDBC: Canberra).)
O'Loughlin, E.M. (1988). Hydrology of changing landscapes. Civil Engineering Transactions. The Institute of Engineers, Australia, CE30, 163-173.
Prime Ministers Science, Engineering and Innovation Council (1999). `Dryland Salinity and its Impact on Rural Industries and the Landscape' Occasional Paper No. 1. (Department of Industry, Science and Resources, Canberra.)
Simpson P. and Langford C. (1996). Managing High Rainfall Native Pastures on a Whole Farm Basis. (NSW Agriculture, Sydney).
Specht R.L. (1972). Water use by perennial evergreen plant communities in Australia and Papua New Guinea. Australian Journal of Botany 20, 273-299.
Whalley R.D.B., Robinson G.G., and Taylor J.A. (1978). General effects of management and grazing by domestic livestock on the rangelands of the Northern Tablelands of New South Wales. Australian Rangelands Journal 1, 174-90.