Facebook Twitter Flickr Google + Blogger

Australasian Plant Conservation

Originally published in Australasian Plant Conservation 16(4), March - May 2008, pp 22-23

Reconstructing complex grassland on agricultural sites by direct seeding: learnings from a 3 year, field-scale, experimental study

Paul Gibson-Roy
The University of Melbourne/Greening Australia (Victoria), c/- Burnley Campus, Richmond. Email: roypg@unimelb.edu.au

Figure 1. Plants grown in polystyrene containers under controlled nursery conditions in one of six seed production areas established to produce seed to supplement field collections. Photo: P. Gibson-Roy

Figure 2. A modified turf seeder allows the sowing of multiple species in one pass. Photo: P. Gibson-Roy

Figure 3. A scalped section within a sowing site (Laharum) 12 months after sowing, with a higher sward of
Austrostipa mollis. A large number of other grass and forb species is also present. Photo: P. Gibson-Roy

Temperate lowland native grasslands are among Australia’s most threatened plant communities (Kirkpatrick et al. 1995). The most common technique for restoring these grassland communities is by the reintroduction of plants germinated from seed and grown as container-stock under nursery conditions. This represents efficiencies in the use of limited seed resources. However, introducing individual plants is labour-intensive and expensive. In addition, the establishment and the development of selfperpetuating populations are seldom reported. Another approach to grassland restoration is by direct seeding. This technique has long been used in Australia for revegetation with native trees and shrubs, but has more recently been investigated for the reintroduction of herbaceous species (primarily grasses and a few selected forbs) (e.g. Windsor and Clements 2001; Gibson Roy et al. 2004; Cole et al. 2005). Some advantages of direct seeding over container grown (or transplanted) introductions, include the potential to work at large scales, lower labour and material costs, and the establishment of large numbers of plants.

To improve knowledge of the potential of direct seeding as a technique for grassland restoration, Greening Australia (Victoria) in partnership with the University of Melbourne undertook the Grassy Groundcover Research Project (GGRP), a three-year NHT funded study.
The focus of the GGRP was twofold:

  • to investigate reintroduction of multi-species assemblages (representative of locally occurring remnants) onto land with an agricultural history; and
  • to investigate the production of large quantities of high quality, provenance seed, in containerised production systems.

Field Methods

The GGRP was initiated in November 2004. Since then three annual sowings have been undertaken at each of 13 one hectare experimental sites across south-western Victoria (39 separate sowings in total). Sowing locations ranged from Bendigo in the state’s central region, to Colac in the south, and from Hamilton in the west, to Minyip in the north. Each of the 13 sowing sites was located on agricultural land with a history of cropping or grazing.

The seed used in the sowings originated from either local grassland remnants (therefore taking into account provenance), or from plants propagated from the wild seed under controlled production facilities (Fig. 1). Most of the 200+ grassland species grown in these production facilities propagated readily from seed and were well suited to the intensive, above-ground seed production system used. Seed production (rather than wild collection) simplified seed harvest and produced reliable quantities of high quality, weed-free seed at times when field production was severely restricted from the effects of drought.

What Has Been Achieved?

At a time when climatic conditions (i.e. drought) were amongst the most challenging in living history, the GGRP sowings have established 13 ha of species rich grassland. Most encouragingly, our results suggest that it is possible to use mechanised equipment (Fig. 2) to sow complex combinations of grassland species under field conditions, and achieve high levels of establishment from a large number of grassland species (Fig. 3). Among these are many locally, regionally and also nationally threatened species. We have also noted high levels of subsequent recruitment of second generation individuals from many species established in each of the first two annual sowings.

Data from the first two years of experimentation suggest that soil removal or ‘scalping’ (Fig. 4) resulted in more effective weed control than either one or two years of traditional chemical control programs prior to seeding. Results comparing three years of chemical control (pre-seeding) with soil removal are yet to be analysed.

Invertebrate surveys have also been initiated to examine colonisation within these reconstructed grasslands. Preliminary results indicate quite rapid utilisation of the sites by a range of native invertebrate species. Investigations of plant roots from individuals established at these sites (across a number of species) indicate the presence of structures associated with functioning arbuscular mycorrhizae. Both findings suggest an increase in functionality at other trophic levels within these reinstated grasslands.

It is hoped that in future the findings from the GGRP will encourage restorationists and land managers to consider the use of direct seeding as a means to increase the range and diversity of herbaceous plant communities.


Cole, I., Lunt, I.D. and 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 southeastern Australia. Restoration Ecology 13: 552-61.

Gibson-Roy, P.G., Delpratt, C.J. and Moore, G.M. (2004). Factors affecting the success of direct-sown native grassland communities. In S.W. Adkins, P.J. Ainsley, S.B. Bellairs, D.J. Coates and L.C. Bell (eds), Native Seed Biology, Australian Centre for Minerals Extension and Research, Brisbane, pp 207-18.

Kirkpatrick, J., Mc Dougall, K. and Hyde, M. (1995). Australia’s most threatened ecosystem: the southeastern lowland native grasslands. Surrey Beatty & Sons, The World Wide Fund for Nature.

Windsor, D.M. and Clements, A. (2001). A germination and establishment field trial of Themeda australis (Kangaroo Grass) for mine site restoration in the Central.