Charles Sturt University
Charles Sturt University


Restoring a kangaroo grass understorey

By Ian Cole, Terrestrial Ecologist, CNR, DLWC, Cowra

Whilst scattered box trees are a relatively common sight across the wheat sheep belt, the diverse plant communities that used to live beneath them have long since gone, replaced with a few agricultural species, weeds and hardy native plants. Animal species that depended on the original plant species for habitat are also largely absent.

White Box WoodlandThese original grass-dominated understoreys are therefore extremely important because they largely determine the aesthetic and conservation value of these communities. In the original plant communities these grasses exerted a powerful influence over other plant species, at some times suppressing species and at other times allowing them to recruit, producing distinctive and diverse plant communities. Restoration of these communities therefore must include strategies that will ensure the re-establishment of the original grassy understorey species. If this can be accomplished other species may be reintroduced at a later stage to restore to the vegetation a structure, function and habitat value to something more like the original.

Right: Fig. 1. This Grassy White Box woodland with severely depleted native understorey is situated at the Centre for Natural Resources (Cowra), and is typical of much of the well farmed wheat sheep belt across NSW and northern Victoria.

Surprisingly therefore, whilst the conditions necessary for the regeneration of the dominant eucalypt component of some of these communities has been well studied, little work has been done on restoring their grassy understorey. Fencing off areas, although successful in allowing regeneration of tree and shrub species, does not necessarily result in regeneration of the original grass species. A more interventionist approach of introducing seed onto the site may be needed in some situations.

Themeda triandra (Kangaroo grass) was one of the more widespread grassy understorey dominants. This resilient grass is well adapted to the variable Australian climate as it can utilise rainfall and soil nutrients efficiently. Its roots enhance soil health and structure by binding and aggregating soil particles, increasing soil porosity, cycling nutrients, increasing soil organic matter and supporting a large number of soil organisms.

Kangaroo grass seed comprises a dark brown caryopsis tightly enclosed by 2 glumes. At one end of this ‘seed’ is a sharp pointed callus while at the other is a long bent awn. Under natural conditions the awn moves the seed along the ground until it finds a crack or stone. A dormancy period between 2-24 months prevents most seed from germinating immediately, and most seed will germinate the following spring as soil temperatures rise to between 20 and 25 degrees celsius.

Harvesting seed

Although seed yield per hectare is low, seed is relatively easy to harvest on a small scale. Readiness for harvest is indicated by a colour change in the plant from a dark green/black to a dark green/red colour. When ripe, seed is shed quickly from the plant and in hot dry windy weather nearly all seed will be lost within a week. By the time the plant has turned red most of the seed has been shed. On the small scale, reasonable amounts of seed can be harvested by cutting and tying seed heads into sheaves and hanging them upside down over a plastic sheet. After about a week nearly all of the seed will have fallen out, together with a bit of chaffy material, to produce a relatively concentrated seed product, which is easily stored. It is best not to remove too much trash, as it will prevent the awns from tangling and matting the seed together. A more innovative method involves dragging a sheet of hessian material over the ripe seedheads. Ripe seed lodges in the hessian, which can then be cut into appropriate sized pieces for sowing.

Restoration TrialSeed-hay can be harvested from moderately large areas by using a whipper-snipper, and raking and packing stems into wool bales. On flatter ground, mowers and balers make harvesting easier work. However, with all seed-hay methods, seed loss in hot windy weather loss can be relatively high. Because the seed content of seed-hay is relatively low, a large amount of product has to be dried and stored. It is probably more cost effective to spread it directly onto the site to be sown after harvest. This allows the seed to fall and find suitable microsites using natural mechanisms. If direct spreading is envisaged, it is not necessary to dry the seed-hay and a forage harvester may be more efficient in harvesting seed.

Right: Fig. 2. Trials at Cowra are being used to develop restoration technologies to reintroduce native understorey species such as kangaroo grass (Themeda triandra) and poa tussock (Poa sieberiana).

The most convenient way of harvesting kangaroo grass seed is to use a brush harvester such as the ‘Grass Hopper’. In a reasonable season this machine is capable of harvesting up to 100 kg of high quality seed-floret material per hour.

The most convenient way of harvesting kangaroo grass seed is to use a brush harvester such as the ‘Grass Hopper’. In a reasonable season this machine is capable of harvesting up to 100 kg of high quality seed-floret material per hour. Seed floret material is a more concentrated product than seed-hay and is easily dried and stored. The floret material bulks out the awned seed preventing tangling and is easily spread out at sowing time. Direct heading kangaroo grass is moderately successful; but lowers seed germination and damages awns, limiting the seed’s self-planting abilities.

For kangaroo grass, producing a pure awned seed product is not particularly useful because the awns tangle and mat to produce a solid mass of seed that cannot be teased out for spreading without damaging the awns. Pure kangaroo grass seed will therefore usually be awnless and may even have the callus removed. Removal of the awn and callus has no marked effect on germination. Although removal of the glumes is possible, and has the positive effect of reducing dormancy, it is difficult to accomplish without damaging the seed; so it is not usually attempted. Because pure kangaroo grass seed has no natural mechanism to place it in suitable microsites, it needs to be sown using conventional planting machinery. Limited amounts of kangaroo grass seed are available commercially at prices ranging from $25 to $600 per kg depending on the type of product and the amount of viable seed contained per kilogram.


Products containing awned seed are probably more suited to direct spreading on inaccessible or non-arable areas than are un-awned seed products. Although awned seed has a remarkable ability to find suitable microsites, establishment rates are poor in areas with complete groundcover of other species or where the soil surface is compacted or crusted. Rough cultivation before spreading seed-hay will therefore improve establishment. Establishment has been improved in some situations by reducing the vegetative cover in spring. A knockdown herbicide is applied in early spring before germination of the kangaroo grass, and is followed by burning the dead material. Alternatively, crash grazing followed by herbicide application produces a litter ‘mulch’ that has proved beneficial to establishment especially in dry seasons.

Pure (un-awned) kangaroo seed may also be sown into a freshly worked but rough seedbed that will allow seed to fall into depressions where it will be covered by loose particles of soil and trash. On non-arable sites where this is not possible, hydroseeding or hydromulching techniques should be considered, but are probably only relevant to small scale well-funded revegetation projects. Apart from this, the better use for pure seed is on relatively flat terrain where it can be accurately sown about 1 cm deep into a relatively uniform seedbed using conventional seeding machinery.

Establishment rates using all these techniques will vary considerably depending on site conditions and season. As a guide, it is reasonable to expect about 5% of spring sown viable seed to develop into established plants by the end of summer from a relatively dry season, and up to 50% in a more favourable season. Even low plant establishment should be considered successful, as kangaroo grass competes well with most species especially where site fertility is low. Even densities of 1-2 plants per square metre are capable of developing into thick stands within 3-5 years, given a couple of favourable seasons. On higher fertility sites the ability of kangaroo grass to compete with other grassy perennials, such as phalaris and paspalum, is much reduced. These perennials should be controlled either by herbicide or cultivation before sowing kangaroo grass.


After the first 6 weeks, young kangaroo grass seems to tolerate light grazing from several types of herbivores including rabbits, deer and kangaroos, although plant numbers are reduced. Establishment areas should, however, be fenced to exclude sheep and cattle in the first year. Once kangaroo grass is established, high intensity-intermittent grazing by sheep or cattle should be timed to suppress weed growth and to prevent the kangaroo grass from becoming too tall and rank. Tall, rank kangaroo grass stands become poorly rooted and are severely damaged by grazing.

Weed suppression can also be augmented by herbicides if this is appropriate to the site. Kangaroo grass seems to be tolerant of Simazine, Atrazine, Verdict®, Glean®, and most selective broadleaf herbicides when they are used at the rates recommended on the label to control specified weeds in pasture.

This article was first published in Woodland Wanderings, 2003.

Further information:
  • Ian Cole, Iain Dawson, Warren Mortlock, and Susan Winder. (2000). Using Native Grass Seed in Revegetation, FloraBank Guideline 9. FloraBank, Canberra
  • Cole I & Lunt ID (2005). Restoring kangaroo grass (Themeda triandra) to grassland and woodland understoreys: a review of establishment requirements and restoration exercises in SE Australia. Ecological Management and Restoration 6(1), 28-33.
  • Cole I, Lunt ID & Koen T (2005). Effects of sowing treatment and landscape position on establishment of the perennial tussock grass Themeda triandra (Poaceae) in degraded Eucalyptus woodlands in south-eastern Australia. Restoration Ecology 13(3), 552-561.
  • Cole I, Lunt ID & Koen T (2004). Effects of soil disturbance, weed control and mulch treatments on establishment of Themeda triandra (Poaceae) in a degraded white box (Eucalyptus albens) woodland in central New South Wales. Australian Journal of Botany 52(5), 629-637.