Facebook Twitter Flickr Google + Blogger

Australasian Plant Conservation

Originally published in Australasian Plant Conservation 17(1) June - July 2008, pp 22-23

Kangaroo Grass: a keystone species for restoring weed-invaded temperate grassy woodlands 

Suzanne Prober1 and Ian Lunt2 
1CSIRO Sustainable Ecosystems, Wembley, WA. Email: suzanne.prober@csiro.au 
2Institute for Land, Water & Society, Charles Sturt University, Albury, NSW. 

Conventional approaches to controlling environmental weeds often target the invading species directly. For example, weeds can be removed manually or killed by herbicides. However, these 'top-down' approaches are often ineffective, because the weed can easily re-invade or be replaced by other weed species. A solution to this problem is to combine 'top-down' weed control with an ecological approach that targets the 'bottom-up' restoration of the native ecological community (Sheley and Krueger-Mangold 2003). 

The key to 'bottom-up' restoration is to recreate an ecosystem that maintains an environment that is unfavourable to weeds, and so is 'resistant' to weed invasion. The key to such ecological resistance is often related to interactions between native species and their environment. For example, a native species may pre-empt resources such as light, nutrients or water, so that weeds are unable to grow, or might secrete allelopathic chemicals that inhibit weeds. Some weeds can invade natural, undisturbed ecological communities, and this approach will not be successful for controlling such species. Commonly though, weeds invade most vigorously when external disturbances destroy the ecological resistance of the native community. 

Restoring Resistance to Weed Invasion 

The 'bottom-up' approach is relevant to restoration of Kangaroo Grass (Themeda triandra) dominated grasslands and grassy woodlands across southern Australia. As a result of livestock grazing, fertilisation and other disturbances, Kangaroo Grass has been lost from many remnants of these ecosystems, and they have instead become dominated by annual weed species and other natives. The weeds outcompete native forbs and change habitat conditions for native fauna, leading to reduced woodland diversity over broad scales. 

How then do we restore natural resistance to invasion by exotic annuals in these grassy ecosystems? What are the ecological processes we need to restore, and what native species drive these processes? 

In an earlier issue of Australasian Plant Conservation (Prober et al. 2004) we described some novel approaches for restoring weed-invaded native understoreys in temperate Kangaroo Grass ecosystems. We showed that areas heavily invaded by exotic annuals had high soil nitrate levels, whereas areas with few exotics had very low soil nitrate levels. We hypothesized that soil nitrate was the key underlying driver of natural resistance to weed invasion in these ecosystems, and that if soil nitrate could be reduced, weeds would grow less and establishment of native plants would be enhanced. We tested these hypotheses in field trials near Young, NSW, and found that in weedy areas with high soil nitrate, Kangaroo Grass was unable to re-establish even when we added lots of its seed (Fig. 1). However, when we actively intervened to suppress weeds, Kangaroo Grass was able to re-establish successfully. 

Figure 1. In this weedy site Kangaroo Grass was not able to establish when we added Kangaroo Grass seed (Control + Kangaroo Grass). However when we suppressed the soil nitrate supply to weeds by adding carbon (sugar), Kangaroo Grass established very effectively (Sugar + Kangaroo Grass). Few weeds were then able to persist in the re-established Kangaroo Grass sward, even when we stopped artificially suppressing nitrate using sugar. Photos: S. Prober Control + Kangaroo Grass Sugar + Kangaroo Grass

We effectively reduced weeds using two very different techniques: (1) by reducing soil nitrate levels using carbon (sugar) additions, and (2) by reducing weed seed banks by burning the standing weed crop before it set seed. Other techniques such as spraying might be similarly effective, so long as Kangaroo Grass seed is added. The first method conclusively demonstrated the importance of soil nitrate: by reducing it, weed growth was greatly reduced and native plant establishment was enhanced. But how sustainable is this approach? The critical next step was to ensure that weeds did not re-invade after we stopped actively controlling them. For example, sugar addition only provides temporary weed control, because soil nitrate increases again after about three months. Similarly, annual weed seeds can move in from surrounding areas even if their seed banks are controlled on-site. 

Three years after we successfully established Kangaroo Grass by adding sugar or burning, we found some exciting results (Prober and Lunt 2008). In many places, the Kangaroo Grass seedlings had grown to form a dense sward (Fig. 1). When we measured soil nitrate levels beneath these swards, we discovered that they had become extremely low, even on burnt plots where we had never added sugar (Fig. 2). Indeed, soil nitrate levels were similar to those we have measured in weed-free, undisturbed reference sites. 

In these places, few weeds were able to grow with the Kangaroo Grass. In more intact remnants, we've found that even when Kangaroo Grass swards are burnt or mown, allowing plenty of light through the recovering sward, nitrogen-loving weeds remain inhibited. It appears then that the ability of Kangaroo Grass plants to lock up soil nitrate is an important mechanism for inhibiting weeds. This has led us to conclude that Kangaroo Grass is a keystone species in these ecosystems, providing long term resistance to invasion by nitrogen-loving exotic annuals through its ability to control soil nitrate. 

Other Native Species 

With Ian Cole from the NSW Department of Environment and Climate Change, we are undertaking further studies to test whether other native grasses are as effective as Kangaroo Grass for providing this ecological resistance to weed invasion in temperate grassy ecosystems. Preliminary indications are that other species are not as effective, suggesting that Kangaroo Grass itself is one of our most important tools for restoring these ecosystems. Another question is whether desirable native species are inhibited by Kangaroo Grass due to its effects on soil nitrate. A study by one of our honours students, Lisa Smallbone, provided illuminating results. Lisa's pot trials indicated that increased levels of soil nitrate can increase the growth rate of native forbs, but that nitrogen-loving exotic annuals are far more inhibited by low soil nitrate levels than are many native forbs. This suggests that while low soil nitrate levels might reduce the growth rate of native forbs, the forbs may still be better off because they don't need to compete with weeds (Smallbone et al. 2008). When Lisa tested this in the field experiment, she found that native forbs established significantly better on plots where we reduced soil nitrate levels to control weeds (Smallbone et al. 2007). 

Figure 2. These data show how soil nitrate was suppressed on plots with established Kangaroo Grass (Themeda) swards (with >75% cover as indicated), compared with higher soil nitrate on plots where Kangaroo Grass didn’t establish successfully (0-3% cover). Soil nitrate on unseeded sugar plots remained somewhat suppressed due to the residual influence of sugar, but was still higher than on the seeded plots. The reference nitrate level indicates levels we typically measure in weed-free reference sites. Different letters indicate that treatments are significantly different at P<0.005. Modified from Prober and Lunt (2008).


Our simple message here is that to restore weed-invaded native understoreys in temperate Kangaroo Grass ecosystems, we need first to focus on restoring swards of Kangaroo Grass, using any effective short-term intervention (such as those we've mentioned). Once we achieve a healthy sward, this species will then do the ongoing weed control for us, by locking up soil nitrate and creating an environment unfavourable to weeds. 

This provides an effective example of a 'bottom-up' approach to weed control. By re-establishing Kangaroo Grass, we remove an underlying driver of weed invasion and create an ecosystem that is resistant to annual weeds in the longer term. 


This work was funded by the NSW government through its Environmental Trust. We thank the Johnson family for generously assisting with these studies on their property 'Windermere' near Young, NSW. 


Prober, S.M. and Lunt, I. (2008). Restoration of Themeda australis swards suppresses soil nitrate and enhances ecological resistance to invasion by exotic annuals. Biological Invasions. DOI 10.1007/s10530-008-9222-5. 

Prober, S.M., Thiele, K. and Lunt, I. (2004). A sweet recipe for understorey restoration in grassy woodlands - add sugar, seed and burn in spring! Australasian Plant Conservation 13: 4-7. 

Sheley, R.L. and Krueger-Mangold, J. (2003). Principles for restoring invasive plant-infested rangeland. Weed Science 51: 260-265. 

Smallbone, L., Lunt, I.D. and Prober, S.M. (2008). Soil nitrate promotes an exotic grass more than native forbs. Ecological Management and Restoration 9: 62-66. 

Smallbone, L., Prober, S.M. and Lunt, I.D. (2007). Restoration treatments enhance early establishment of native forbs in a degraded grassy woodland. Australian Journal of Botany 55: 818-830.