module 8

Rangeland weeds

Objectives

This module is designed to bring an understanding of the management of weeds in Australia’s rangelands. At the end of this topic you will understand:

Module Outline

Introduction

(Notes based on: Grice, AC 2000, ‘Weed management in Australia’s rangelands’, in Sindel, BM (ed.) Australian Weed Management Systems, RG and FJ Richardson, Melbourne, Victoria, pp. 431-58.)

Over half of Australia’s land area is used for grazing livestock on plant communities typically denoted as ‘grasslands’ or ‘pastures’. Such pastures have often been described as native, natural or improved.

Another classification divides these grasslands into ‘rangeland’ communities, which are typically comprised of native plants in arid, vast, open and sparingly fenced areas and ‘pasture’ communities, predominantly of introduced species, whether sown or volunteer.

In this topic the focus is on weeds of rangeland communities.

Rangeland communities

Rangelands have been defined as ‘semi-natural eco-systems in which people seek to obtain a productive output by simply adding domestic stock to a natural landscape’.

This definition (Harrington et al. 1984) can be applied to about 70% of the Australian mainland, an area of around 6 million km2 that had a total human population of <500,000 in 1990 (Foran et al. 1990).

Australia’s rangelands have very diverse climates which range from arid to semi-arid and temperate to tropical. Amount and timing of rainfall is variable, but is usually more summer-dominant in the north and winter-dominant in the south.

Many soils of Australian rangelands are infertile and prone to erosion, though there are exceptions such as the self-mulching, cracking clays that support Mitchell grass (Astrebla spp.) communities in central north Queensland.

Vegetation types include chenopod shrublands, tussock grasslands, hummock grasslands, mallee, semi-arid, arid, tropical and sub-tropical woodlands and temperate rangelands of the south-east (Harrington et al. 1984). About 63% of Australian rangelands are used for pastoralism. Other uses are for conservation reserves, military training and dedicated aboriginal lands.

Effective weed management in these communities involves knowledge of the types of weeds present in the rangeland, and their ecological role and impact on the native plant communities. It also requires an understanding of the characteristics of weed management, including the objectives, concepts and principles and the strategies and practices relevant to the effective management of weeds in these rangelands.

Weeds of Australian rangelands

Most plant species in Australian rangelands are indigenous, but many others are alien. From an environmental perspective, all alien species could be regarded as weeds.

Of alien species identified in Central Australia and Western NSW, approximately 25% have been declared noxious. Noxious weeds are mostly annual forbs (>50%), with the rest being perennial forbs, annual grasses, shrubs, trees and perennial grasses.

A few indigenous rangeland species have also been classed as noxious in parts of Australia. Some examples are:

Many indigenous species are perceived as weeds in various situations. At least 50 indigenous species, including grasses, forbs, shrubs and trees, from western NSW, have been identified as toxic to livestock.

One example of an indigenous noxious weed is Ixiolaena brevicompta (flat billy button). This native weed has been responsible for heavy losses of livestock in floodplain country.

Woody weeds

Certain indigenous shrubs are widely recognised as weeds in various areas of the rangelands. They are ‘native woody weeds’, and are especially prominent in the semi-arid woodlands of western NSW and south-west Queensland. These species have not been declared noxious, but have been formally recognised as ‘woody weeds’.

Some examples of native ‘woody weeds’ are:

In northern Australia there have been increases in indigenous shrubs and trees in what were thought to be previously open communities. However, alien species are perhaps more prominent here than in the southern rangelands.

Most serious alien species are:

Each has been declared noxious in Queensland, Northern Territory and/or Western Australia.

Herbaceous weeds

A large number of herbaceous alien species are invasive in particular rangeland types, although there is often conflicting opinion due to differing land use objectives.

Many perennial grasses and legumes have been tested for their forage potential. A large proportion of these have subsequently become weeds in rangelands and agricultural areas, or in other areas because of their impact on those eco-systems and their conservation values.

One example is Cenchrus ciliaris (buffel grass). This persistent perennial grass invades areas in central Australia, an area which is crucial for flora and fauna conservation.

Other examples are the grasses Andropogon gayanus (gamba grass) and Urochloa mosambicensis (sabi grass), and the legumes Stylosanthes hamata (Caribbean stylo), S. humilis (Townsville stylo), and S. scabra (shrubby stylo).

Impact of weeds on rangelands

There are several problems associated with ‘woody weeds’ – native or alien. The woody weeds of most concern are those which impact on the structure and function of eco-systems, whether for conservation or pastoral reasons.

Increased woody plant cover has been linked to declining herbaceous yield. Through their effects on forage quantity and quality, shrubs are thought to lead to increased mortality of animals, as some plants have been proven to be toxic to native animals and to livestock, and to a reduction in overall livestock carrying capacity.

Increased shrub cover may inhibit mustering of livestock and reduce animal movement. This may make it more difficult for animals to gain access to water and may also reduce contact between animals at joining time and lead to mis-mothering of lambs.

Dense populations may harbour feral animals.

Herbaceous rangeland weeds have a variety of impacts. Some compete with more desirable pasture species and some are toxic to livestock and contribute to animal health problems.

The awned seeds of various native grasses e.g. Aristida spp. (wiregrass) and Stipa spp. (speargrass) and introduced grasses, e.g. Hordeum spp., contribute substantially to vegetable contamination of wool and cause animal health problems when large numbers penetrate the eyes, mouth and flesh of grazing stock.

Economic impact of rangeland weeds

Estimates indicate that the areas of rangelands infested with weeds are large. In western NSW, an estimated 20 million ha were actually or potentially affected by the proliferation of indigenous woody plants in 1983 (Alchin et al. 1979).

Large areas of rangelands are also affected by alien woody shrubs. For example, in Queensland, rubber vine was affecting an estimated 3 million ha in 1990 (McFadyen and Harvey 1990), while prickly acacia probably covered over 7 million ha in 1997 (Mackey 1997).

Declining carrying capacities have been attributed, at least partly, to the proliferation of woody species and alien shrubs. Shrub encroachment was seen as a problem by 50% of landholders surveyed (Hassall and Associates 1982).

Environmental effects of rangeland weeds

Significant effects have been attributed to rangeland weeds, but these have not been accurately measured. Several species dramatically alter the structure and composition of rangeland plant communities that they invade.

Characteristics of weed management in rangelands

Several factors severely constrain weed management in rangelands.

Firstly, the areas involved are vast, but plant production and economic returns per unit area are minimal. Because of this vastness, resources are generally limited relative to the extent of the problem. Extensive application of relatively expensive chemical and mechanical techniques is difficult to justify.

Secondly, the density of the human population is very low. This means there are few human resources available for weed management.

Finally, the weeds occur in complex and dynamic plant communities that are subject to multiple land uses. Often there is more than one ‘problem weed’ in a situation, so management approaches must deal with ‘weed complexes’ rather than individual species.

Objectives of weed management

The objectives of weed management in Australian rangelands, and the concepts and principles on which practice is based should reflect the underlying constraints. Once a weed has become established eradication is highly improbable. Containing and reducing the impact of existing populations are more realistic objectives.

There are several important ecological concepts which can be linked with the principles of weed management that are appropriate for Australian rangelands. They include:

Some important ecological concepts

Dispersal

Limiting dispersal can help contain a weed’s infestation. Knowledge of dispersal mechanisms may show how spread can be minimised and knowledge of a weed’s phenology may indicate the most suitable time for action.

For example, the spread of prickly acacia can be minimised by not transporting cattle that have recently had access to pods containing viable seeds (Brown and Carter 1998).

Competition

Interspecific plant competition is one factor governing the composition of plant communities. Reduced competition from perennial grasses as a result of heavy grazing has encouraged the proliferation of native woody species in grassland and woodland systems.

One experiment (Harrington 1991) demonstrated an increased survival of seedlings of narrow-leaved hopbush when perennial grasses were removed.

Disturbance

Disturbance of an ecosystem has been indicated as a pre-requisite to invasions by exotic plants. However, not all types of disturbance have the same impact, and some types or degree of disturbance occur in all ecosystems.

The invasion by lantana (Lantana camara) of dry rainforest in North Queensland has been correlated with soil disturbance by feral pigs, increased light levels that result from consequent tree death, and increased intensity of fires as a result of changed fuel characteristics.

Invasion of tropical woodlands by Indian couch has been associated with the disturbance of herbaceous vegetation by heavy cattle grazing.

Fire can be a means of disturbance. However, fire can also be a means for countering invasion by rubber vine and also for increasing the abundance of pastorally desirable native grasses such as Mitchell grass over less desirable species.

Spatial patterns of invasion

Like most plants, weeds are usually spread unevenly across the landscapes in which they occur. Invasive plants favour some habitats over others (e.g. rubber vine does best in riparian habitats and prickly acacia thrives along bore drains).

The distribution of weeds also reflects historical factors such as the locations of sites at which they were first introduced.

In most cases eradication is not feasible; therefore knowledge of the spatial processes of invasion will help ensure that limited resources are directed at populations which pose the greatest threats.

Temporal patterns of change

In many Australian rangelands, rainfall is erratic and water is often a limiting factor for plant growth. Australia’s frequent drought periods favour weeds that have developed long-lived seedbanks or other survival mechanisms.

Models have been proposed that describe the risks and opportunities associated with episodic events such as drought or very wet periods. These models present the idea that vegetation change in many rangelands, including changes in the abundance of native woody weeds and invasive exotic shrubs and grasses, is often better described in terms of episodic events rather than a continuous process.

In the Cobar/Byrock region of NSW, three major waves of seedling establishment of native shrubs in semi-arid woodlands occurred during exceptionally wet periods – the 1890’s, 1950’s and 1970’s.

The episodic increase of prickly acacia on Mitchell grasslands in western Queensland apparently occurred when a shift from sheep to cattle grazing coincided with a run of years of high rainfall (Brown and Carter 1998).

icon
Reading

‘Weed ecology and population dynamics’ (pages 46 to 57) in Australian Weed Management Systems.

Principles for rangeland weed management

These basic ecological concepts can be linked to principles for the management of weeds in Australian rangelands. These principles include:

Rangeland weed management systems

The preceding objectives, concepts and principles for rangeland weed management provide a common-sense approach to developing weed management systems.

Some of them, including the need to prevent, detect and contain infestations and to take integrative and strategic approaches, have been incorporated into the Australian National Weeds Strategy (Commonwealth of Australia 1997).

Prevention

The need to prevent the introduction of new weeds to Australia is an objective of the National Weeds Strategy. The strategy also recognises the need to prevent the spread of weeds to new areas within Australia.

The national strategy for the control of prickly acacia aims to eradicate scattered infestations and prevent the plant’s establishment in areas where it is currently absent.

The main area of infestation is on the Mitchell grasslands in central Queensland. The Mitchell grasslands are where the species was widely planted as a fodder and shade tree.

Learning activity

Go to Activity 8-1

Figure 8.1

Figure 8.1 Distribution of Acacia nilotica (prickly acacia) in Australia showing major infestations (shaded) in the Mitchell grass zone in Queensland and scattered infestations (X) elsewhere (after Kriticos 1997).

For several important rangeland species control of animal-aided dispersal is an option.

Domestic cattle are major dispersal agents of invasive shrubs such as prickly acacia, mesquite and chinee apple.

Prediction

Bioclimatic analysis, involving computer software programmes such as CLIMEX and BIOCLIM, has been used to predict the potential ranges of invading species. Range predictions for species relevant to Australian rangelands include those for rubber vine

(Tomley 1995) and prickly acacia (Mackey 1997). There have been no tests of the reliability of these predictions, because this needs data independent of that used to develop the predictions. Accumulation of such data could take years or even decades to complete.

Little attention has been given to predicting landscape scale patterns of spread. This would be potentially the most useful from a practical weed management point of view because of the need to focus on containment and minimising impacts.

Predictions have been made for a few species at a regional level, e.g. Nassella trichotoma (serrated tussock), Opuntia aurantiaca (tiger pear) and Parthenium hysterophorus (parthenium weed), but have not been quantified at landscape scales.

Prickly acacia is one species relevant to Australian rangelands for which some effort has been put into predicting landscape scale spatial dynamics. This species grows more densely along creek lines and bore drains than it does in adjacent uplands. Populations along bore drains may be controlled using herbicides, and capping bores, in which case, upland populations may become unsustainable.

Rubber vine is an invasive shrubby vine introduced during the late 19th Century. Its range now covers an area of 350,000 km2, but although the species is tolerant of a wide range of soil types and climates in the wet and dry tropics, it is not evenly spread across the region.

At the landscape scale, rubber vine is more abundant in riparian habitats, most likely because of the high soil moisture levels found there. In drier areas, the species is more or less restricted to riparian zones. An understanding of these aspects would facilitate qualitative predictions of the spatial pattern of invasion.

Detection

Some systematic efforts have been made to detect specific weeds in rangelands. A combination of public reports, field and aerial surveys, aerial photographs and satellite imagery has been used to detect giant sensitive plant in the Northern Territory. The results of this work have been used to target reasonably large-scale, publicly-funded control programmes.

Remotely sensed data have been used to detect trends in woody plant biomass in areas prone to increase in native shrubs. This type of information is more useful for describing general trends and highlighting the scale of a weed problem than for targeting management.

Likewise, attempts have been made to produce distribution maps. Generally, the scale of these maps and the precision of the information from which they were derived are such that they are useful only for considering national scale management strategies.

Most examples of detection of previously unknown weed infestations are probably fortuitous. Rubber vine was assumed to be restricted to Queensland until 1997, when it was discovered growing near the Ord River in the north of Western Australia.

Detection of incipient infestations at scales relevant to regions, catchments and management units will always rely largely on local experience combined with skills in plant identification.

Early intervention

The principle of early intervention has been recognised in connection with various weed problems in Australian rangelands. Control action has been taken against a number of potentially serious weeds while their ranges are still quite restricted.

Tamarix aphylla (Tamarisk, Athel pine) is native to northern Africa and parts of south west Asia. In Australia, naturalised populations are currently restricted to the vicinity of the Finke, Ross and Palmer Rivers in the Northern Territory.

A management strategy for this species aims to:

The strategy focuses on treating upstream populations and preventing further downstream spread of the seeds (which are thought to be dispersed downstream in floodwaters).

Kochia scoparia (kochia) (imported in 1990) was widely sown on salt-affected land in Western Australia as a salt-tolerant forage plant for pastoral and cropping areas. Within two years its weed potential was recognised and it was targeted successfully for containment but less successfully for eradication.

The success of this programme was facilitated by intervention within two years of initial introduction, the ability to trace most of the sites of introduction and a well-organised government-funded campaign.

In most situations, intervention has rarely been early enough for eradication to be a realistic goal, but early action against particular infestations would help reduce local impacts (e.g. the use of fire to prevent the spread of native shrubs in NSW and Queensland).

Dense shrub populations suppress grasses and reduce the probability of fire to the point where it is no longer a management option. Burning before shrub populations start to suppress the grasses would constitute early intervention in this system.

Integrated management

Integrated approaches to weed management have been advocated for rangelands, however research has focused on examining individual techniques for controlling woody species.

The four main techniques promoted for control of woody plants in rangelands are: mechanical treatment, use of herbicides, burning and biological control. Some attention has also been given to the use of browsing herbivores (e.g. goats) to control native and exotic shrubs.

Even though integrated management is promoted in principle, there are few examples where the major techniques have been integrated into well-developed control programmes. The most obvious form of integrated management is the sequential use of two or more techniques against a single infestation

One example is the use of chemical defoliants as a follow-up treatment to the use of fire for the management of sprouting native shrubs such as budda, turpentine and punty bush, but the approach has not been used on a commercial basis.

Several methods are available for the control of rubber vine:

The possibilities involving various combinations of these techniques have not been fully explored, for example plants surviving mechanical treatments could be treated with herbicides (Tomley 1995), and defoliation by the biological control agents may allow grass growth that in turn facilitates more effective fires (Tomley and Hardwick 1996).

Strategic management

Integrated management requires combinations of control measures that are effective against individual weeds or suites of weed species. Strategic management involves applying these measures at times and places that gain the greatest benefit for the costs involved (e.g. for the control of rubber vine).

The use of buffer zones is a strategy designed to prevent the spread of rubber vine into the Northern Territory and involves the use of two zones, a western one from which the species is to be eradicated, and an ‘active control zone’ further east.

Within individual catchments, the strategy proposes to begin control activities at the head of the catchment and work downstream, based on the idea that rubber vine is more likely to spread downstream. Co-ordination and funding are important for national and state-wide problems.

Benefits are likely to accrue when groups of land managers, such as Landcare groups, co-ordinate their activities (e.g. where an infestation straddles property boundaries). Group action should concentrate on those infestations that pose the greatest threat to habitat.

Strategies proposed for the management of native shrubs involve ranking infestations from high to low priority:

Effective weed management

Comprehensive weed management systems are emerging for some of the exotic, invasive weeds of Australia’s rangelands. An effective system for these species will entail most of the following features:

Learning activity

Go to Activity 8-2

Conclusion

Rangeland weeds include both indigenous and alien species which can affect the structure and function of eco-systems and pastoral communities. Weed management in Australian rangelands is constrained by extensive areas of threatened land, low density of human population and low economic returns. Prevention and containment may be achieved by understanding the key ecological processes of dispersal, competition, disturbance, and spatial and temporal patterns of vegetation change.

The principles of weed management that include prevention, prediction, detection and early intervention, form the basis for effective weed management strategies.

References

Alchin, BM, Proude, CK and Condon, RW 1979, ‘Control of woody weeds in western New South Wales’, Proceedings of the 7th Asian-Pacific Weed Science Society Conference, pp. 141-3.

Amor, RL & Piggin, CM 1977, ‘Factors influencing the establishment and success of exotic plants in Australia’, in D Anderson (ed.), Exotic Species in Australia - their Establishment and Success, Proceedings of the Ecological Society of Australia 10, pp. 15-26.

Brown, JR and Carter, J 1998, Spatial and temporal patterns of exotic shrub invasion in an Australian tropical grassland. Landscape Ecology 13, 93-102.

Commonwealth of Australia 1997, The National weeds Strategy: a Strategic Approach to Weed Problems of National Significance, Commonwealth of Australia, Canberra.

Crawley, MJ 1996, Plant ecology, 2nd edn, Blackwells Scientific Publications, Oxford.

Dowling, PM, Michalk, DL and Sindel, BM 2000, ‘Weed management in pasture systems’, in Sindel, BM (ed.) Australian Weed Management Systems, RG and FJ Richardson, Melbourne, Victoria, pp.. 309-28.

Foran, BD, Friedel, MH, MacLeod, MD, Stafford-Smith, M and Wilson, AD 1990, A policy for the future management of Australia’s rangelands, CSIRO, Canberra.

Grigulis, K, Sheppard, AW, Ash, JE & Groves, RH 2001, ‘The comparative demography of the pasture weed Echium plantagineum between its native and invaded ranges’, Journal of Applied Ecology, vol. 38, pp. 281-290.

Groves, RH & Kaye, PE 1989, ‘Germination and phenology of seven introduced thistle species in southern Australia’, Australian Journal of Botany, vol. 37, pp. 351-359.

Harper, JL 1977, Population Biology of Plants, Academic Press, London.

Harrington, GN 1991, ‘Effects of soil moisture on shrub seedling survival in a semi-arid grassland’, Ecology, vol. 72, pp. 1138-49.

Harrington, GN, Wilson, AD and Young, MD 1984, Management of Australia’s Rangelands, CSIRO, Melbourne.

Hassall and Associates 1982, ‘An economic study of the Western Division of NSW’, A report prepared for the Western Lands Commission. Hassall and Associates Pty Ltd, Canberra.

Kriticos, DJ 1997, ‘A climatic analysis of the potential distribution of prickly acacia in Australia’, Report prepared for Queensland Department of Natural Resources, Brisbane.

Leigh, JH & Holgate, MD 1979, ‘Effects of pasture availability on the composition and quality of the diet selected by sheep grazing native, degenerate and improved pastures in the Upper Shoalhaven Valley, New South Wales’, Australian Journal of Experimental Agriculture and Animal Husbandry, vol. 18, pp. 381-390.

Lonsdale, M 1996, ‘The balance of weed populations’ in RB Floyd, AW Sheppard and PJ DeBarro (eds), Frontiers of Population Ecology, CSIRO Publications, pp. 573-588.

Mackey, AP 1997, ‘The biology of Australian weeds. 29. Acacia nilotica ssp. indica (Benth.) Brenan’, Plant Protection Quarterly vol. 12, pp. 7-17.

Medd, RW, Kemp, DR & Auld, BA 1987, ‘Management of weeds in perennial pastures’, in JL Wheeler, CJ Pearson & GE Roberts (eds), Temperate pastures, their production use and management, CSIRO Publications, pp. 253-261.

McFadyen, RE and Harvey, GJ 1990, ‘Distribution and control of rubber vine, Cryptostegia grandiflora, a major weed in northern Queensland’, Plant Protection Quarterly, vol. 5, pp. 152-5.

Moore, RM 1970, Australian Grasslands, Australian National University Press, Canberra.

Noy-Meir, I, Gutman, M & Kaplan, Y 1989, ‘Responses of Mediterranean grassland plants to grazing and protection’, Journal of Ecology, vol. 77, pp. 290-310.

Piggin, CM & Sheppard, AW 1995, ‘Echium plantagineum L.’ in RH Groves, RCH Shepherd, RG Richardson (eds), The Biology of Australian Weeds, vol 1 RG & FJ Richardson, Frankston, pp. 87-110.

Rowarth, JS, Rolston, MP and Johnson, AA 1990, ‘Weed seed occurrence in ryegrass seedlots’, Proceedings of the 43rd New Zealand Weed and Pest Control Conference, pp. 125-129.

Rowarth, JS, Rolston, MP and Johnson, AA 1991, ‘Weed seed occurrence in cocksfoot seedlots’, Proceedings of the 44th New Zealand Weed and Pest Control Conference, pp. 296-299.

Tomley, AJ 1995, ‘The biology of Australian weeds 26. Cryptostegia grandiflora R.Br.’, Plant Protection Quarterly, vol. 10, pp. 122-30.

Tomley, AJ and Hardwick, G 1996, ‘Bulking up, field distribution and establishment of rubber vine rust, Maravalia cryptostegiae in far north Queensland’, Proceedings of the 11th Australian Weeds Conference, pp. 237-8.

icon

Self assessment

Do you know?

  • the major types of pasture weed and their ecological characteristics
  • the role of competition in pastures, and their processes that lead to weed problems
  • the basics of pasture weed population ecology and why it is important?