Seed inoculation with effective root-nodule bacteria enhances revegetation success

Summary

Much of Australia's native arid and semi-arid habitats have suffered from persistent vegetation clearance resulting in the degradation of remnant plant communities. This has led in some areas to an increase in soil salinity and erosion, a decrease in water quality and agricultural productivity, and a loss in biodiversity. Revegetation is therefore a high priority in these areas but many large-scale approaches involve considerable expense. Poor fertility of soil due to the loss of nitrogen and phosphorous, limits the diversity, structure and long-term viability of plant communities but leguminous species may be able to persist due to their ability to form symbiotic associations with nitrogen-fixing rhizobia. Although this interaction is well recognised for its importance in agriculture, few studies have looked at its potential role in habitat restoration. This research investigates the benefits of using legume-rhizobium symbiotic relationships to revegetate degraded areas.

Study sites: A total of nine sites were incorporated in the study, eight in north-central Victoria and one in New South Wales: Donald, St Arnaud, Newstead, Ravenswood, Shelbourne, Timor West, Alma, Bealiba, Hoskinstown. Climatic conditions were similar at all sites i.e. cool, moist winters and hot, dry summers. Each area had suffered from large-scale clearance of native vegetation as well as the long-term effects of livestock grazing. All had low nitrogen levels and non had ever been fertilised. Field trials were initiated in August and September 2002, a time when south-eastern Australia was in severe drought.

Soil characteristics: At least 25 soil samples (10 cm in depth and 4 cm in diameter), were taken at each site. Measurements of pH, salinity and rhizobia numbers were taken. Large soil rhizobial populations if, present, could cause intense competition with the inoculum introduced to Acacia seeds (see Field trials, below) for nodule formation on the hosts.

Seed inoculation: Rhizobial strains were chosen for their effectiveness of nitrogen-fixation within Acacia (obtained from the CSIRO Plant Industry collection). Acacia seeds were then inoculated in sterile peat with the appropriate rhizobium strain and stored at 15ºC. The viability of these was assessed once a month by counting populations of rhizobia on inoculated seeds. No fertiliser was applied but 2-m wide strips were treated with glyphosate weed killer to reduce competition from other plants with germinating seeds.

Sowing: Each site was seeded with a range of native leguminous and non-leguminous shrubs and trees, including several species of Acacia. At each site, two different seed treatments were sown:

1) Coated, inoculated seed.
2) Uncoated, inunoculated seed.

During sowing, the seeds were sprayed with smoked water in order to aid the breaking of dormancy. Seeds were sown 4 m apart, in parallel adjacent rows with identical sowing rates (average of five seeds/m).

Seedling establishment, growth & survival: A series of site censuses were carried out on permanently marked 25 m sections of each seed row. These were used to map the location of newly emerging seedlings for both treatments at each site. Adjacent sections of the two treatments were used to reduce any confounding effects of other variables such as microclimate and topology. Plant establishment, survival, development and height were recorded.

Counts of rhizobial populations: Rhizobia were counted 1) in the soil at each site,
2) on coated, inoculated seeds and 3) in seedling rhizospheres.

A spread-plate technique was used to determine rhizobial numbers on coated, inoculated seeds. Most probable numbers (MPN) were estimated for rhizobia in soils and seedling rhizospheres, using plant infection and nodulation frequency tests (see original paper).

Effects of drought on germination & herbicde application: At the time of sowing, south-eastern Australia was in severe drought, and although normal rainfall resumed by April 2003, the average precipitation deficit across the sites was 30%. As a result the Acacia seedlings germinated much later than in normal years, by which time the effects of the glyphosate weed killer had worn off and competition from weeds was intense.

Rhizobia counts: The average number of rhizobia on coated, inoculated seeds was consistently high (initially between 4.9 and 5.8 rhizobia per seed depending on the Acacia species) although numbers did decrease slightly in storage. The number of naturally occurring rhizobia at most sites was low immediately before sowing in 2002, and decreased further by June 2003, most likely due to the soil moisture deficit. Despite this, high numbers of rhizobia were found in the rhizospheres of seedlings sown at Ravenswood, Shelbourne and Hoskinstown, indicating that a small populations of rhizobia can still be sufficient for successful rhizosphere colonisation.

Seedling emergence & establishment: The first census in November 2002 indicated high variability in germination rates, again probably largely due to the soil moisture deficit. By April 2003, many seedlings had establishment, and this census plus those in June 2003 and February 2004, showed significantly greater establishment of treated than untreated seedlings at all sites except for Hoskinstown (see Table 1, attached). There was little spatial variation in establishment within sites, however there was variation in establishment success for individual Acacia species at the different sites.

Seedling survival: Survival of established seedlings was generally greater for inoculated than uninoculated treatments, although percentage survival did vary considerably between sites (0.7% - 31% for inoculated seeds).

Growth of Acacia seedlings: As it was only feasible to collect complete census data for three sites (Donald, St Arnuaud and Ravenswood) seedling growth was analysed only for these sites. Although some species of Acacia showed a significantly greater increase in height when inoculated than uninoculated, there was no overall consistent increase in growth with inoculation. However, collection of accurate growth data was hindered by preferential grazing at some sites.

Conclusions: Legume nodulation is dependent on the establishment of a threshold rhizobia population, and this is likely to be reached more quickly by plants whose seeds have been inoculated with rhizobia. Inoculant rhizobium strains are more effective at nitrogen fixation than naturally occurring soil rhizobium. Large-scale revegetation by seeding has formerly been constrained by the limited availability of seeds of native species. This study highlights the importance of inoculation of native leguminous species with rhizobia in increasing the rate and success of land restoration with minimal input and maintenance.


Note: If using or referring to this published study, please read and quote the original paper. Please do not quote as a www.conservationevidence.com case as this is for previously unpublished work only.