Facebook Twitter Flickr Google + Blogger

Australasian Plant Conservation

Originally published in Australasian Plant Conservation 20(4) March - May 2012, p 7-9

Monitoring a nutrient manipulation experiment to restore grassy box-gum woodlands

Don A. Driscoll
Friends of the Pinnacle, ACT (www.fotpin.org.au). Email: don.driscoll@anu.edu.au

Diagram of one experimental block (of 10). The fenced and unfenced areas, each have five 5 m × 5 m plots and each plot has a different treatment.

Diagram of one experimental block (of 10). The fenced and unfenced areas, each have five 5 m × 5 m plots and each plot has a different treatment. Plots inside and outside the fenced areas were paired based on their plant composition, then treatments were randomly allocated to pairs of plots. The fenced areas are 12 x 18m, with fences built to exclude kangaroos and rabbit

Sugar sprinkling at The Pinnacle. Photo: Correa Driscoll

Sugar sprinkling at The Pinnacle. We apply a total of 250 kg of sugar every four months to our experimental quadrats, at a rate of 0.5 kg/m2, to test if if nutrient management favours native species. Photo: Correa Driscoll

Volunteers work together to identify plants and record the data in a quadrat within a fenced area at The Pinnacle Nature Reserve. Photo: Don Driscoll

Volunteers work together to identify plants and record the data in a quadrat within a fenced area at The Pinnacle Nature Reserve. Photo: Don Driscoll

Weeds and the Nutrient Hypothesis

Exotic grasses and other weeds have wrought the most profound environmental change throughout The Pinnacle Nature Reserve in the ACT and similar degraded grassy box-gum woodlands in south-east Australia. These weeds are a threat to woodland flora and fauna as they cover the majority of the 138 ha Reserve, and exclude most native plants. Traditional weeding methods cannot defeat this weed menace and new techniques are demanded.

Recent research suggests that removing nutrients from the soil may tip the balance back in favour of native species. In small scale experiments, nutrients have been reduced by addition of sugar or activated carbon (Kulmatiski and Beard 2006; Prober et al. 2005), and this resulted in reduced cover of exotics and increased cover of native species. However, adding sugar is very expensive and requires reapplication every few months. It is not really suited to restoring large areas of degraded grassy woodland (Rawlings et al. 2010). However, adding sugar can provide important insight into the potentially limiting role of nutrients, and therefore the benefits expected if nutrients can be reduced using some other, cheaper method.

A Community-driven Experiment

With funding from the Australian Government’s Caring for Our Country, the ACT Government and the Canberra Labor Club, the Friends of the Pinnacle (Fotpin) have launched an experiment to explore methods for reducing exotic weed cover and increasing native species. The two main goals of the project are to discover:

  • if nutrient management does favour native species in our area (by using a sugar treatment), and
  • if so, how we can reduce nutrients over large areas in a cost-effective manner and restore the native grasslands on the site.

Mammalian herbivores are an additional complicating factor. Grazing by rabbits and kangaroos is substantial at the Pinnacle Nature Reserve, evidenced both by the obvious prevalence of these animals (2.1 per ha in Fotpin’s 2011 surveys), and by our observations that grasses have been grazed back to an extent that has limited flowering. If grazers remove plant material but then replace the nutrients through urine and dung, then this may limit our capacity to remove nutrients. Furthermore, heavy grazing by herbivores may also limit the capacity of native plants to re-establish in our experimental plots. Before we can implement widespread restoration activities, it is critical to understand how important herbivore management might be.

Experimental Design

In ten sites, we have established fenced and unfenced areas, each with a control plot and four treatments for reducing soil nutrients:

  • sugar (0.5 kg sugar/m2 applied every four months)
  • slash removal (autumn and spring annually)
  • burning (applied with a weed burner in conditions when fires will otherwise not go, annual)
  • sterile crop harvest (slash removal in autumn, then rye-corn seeds applied to soil surface and ‘harvested’ in late spring, annual).

We surveyed each site before applying any of the treatments, and monitor annually in spring.

Monitoring – our citizen science approach

We expected that responses to the treatments could vary in a species-specific way. Although we may end up with higher native plant cover, there is a risk that some native plants may decline under particular treatments. We need to have that information to make informed decisions about possible trade-offs between the magnitude of the benefit (in terms of more natives and less exotics) and the magnitude of any disadvantages (in terms of cost and lost native species). Our monitoring was therefore designed to yield detailed information about species‑specific responses.

We use four 1 m2 quadrats within the 5 x 5 m plots. Our quadrats are made of PVC pipe with strings dividing the square into nine sub-squares (like a noughts and crosses board). Within each quadrat we estimate the percent cover of every species, and count the number of sub-squares that each occurs in. These two measures provide information on overall cover and evenness of distribution within the quadrat. However, more importantly, these two measures provide different quality information. Percentage cover is potentially a very useful detailed measure of the predominance of species in a community. However, it is difficult to estimate accurately, and is dependent on the state of development of the plant. We measure percent cover as the area of the canopy of the plant (imagine a bubble encapsulating the canopies of clusters of the same species); this will vary over time for annual species, and for species with voluminous but ephemeral flowering heads. It is also harder for volunteers to estimate accurately, so we are expecting percent cover data to suffer from a fair amount of measurement variability. In contrast, occurrence within the 9 sub-squares is less vulnerable to the errors and seasonal variation of the percent cover method. It also has the advantage of providing some spatial structure for the search within the 1 m2 quadrat. Such detailed surveys require substantial effort. Community members have volunteered a lot of time, and are building substantial expertise in plant identification. Most volunteers spend two or three half day sessions during the four-weekend survey periods. Experienced people work with new volunteers, and everyone’s work needs to be checked by the project leader to ensure consistency and thoroughness. In spring of 2011 we accumulated 243 hours of effort to complete the surveys and slash the plots.

In addition to the plant surveys we also monitor soil nutrients, including total and available phosphorus and nitrogen. We took samples before implementing any treatments, and will take another sample at the end of the project.

We began monitoring in spring 2010 and with the exception of burning, we began our treatments in 2011. Burning will begin in autumn 2012 if the grass dries out enough. Our spring 2010 data revealed there were no significant differences between our treatments. We can therefore be confident that any effects that emerge over the coming years are caused by our treatments and not by some inadvertent confounding factor. We will continue to monitor the flora annually for three years; a timeframe over which we hope to see some response to our treatments.


Implementing a community-based experiment requires four elements to be successful:

  • a project leader with expertise in experimental design, analysis and field methods, or with access to advice in these matters and the ability to learn quickly
  • a group of committed volunteers willing to get down on hands and knees to survey plants
  • a cooperative and engaged park management agency, and
  • a successful grant application.

If you have the first three, the fourth is not that hard to achieve.

Our community-based park-care group hope that our experiment can help us to discover if nutrient manipulation will reduce exotic grass cover and increase natives. Our group, and many others that work in grassy box-gum woodlands need this information so that we can restore extensive degraded areas of this critically endangered ecological community. By undertaking detailed species-level monitoring we expect that we will be able to understand the full range of responses to our treatments. With these insights, we will be able to make informed decisions based on whether the benefits of particular treatments outweigh the costs.

We will then be able to work towards the restoration of our Reserve knowing that our volunteer effort is targeted in the most effective way to get the best restoration results.


Kulmatiski, A. and Beard K.H. (2006) Activated carbon as a restoration tool: Potential for control of invasive plants in abandoned agricultural fields. Restor. Ecol. 14:251-257

Prober, S.M., Thiele, K.R., Lunt, I.D. and Koen, T.B. (2005) Restoring ecological function in temperate grassy woodlands: manipulating soil nutrients, exotic annuals and native perennial grasses through carbon supplements and spring burns. J. Appl. Ecol. 42:1073‑1085

Rawlings, K., Carr, D. and Freudenberger, D. (2010) Box gum grassy woodlands. A land manager’s guide. Greening Australia, Canberra, Australia