Australasian Plant Conservation
Originally published in Australasian Plant Conservation 20(3) December 2011 - February 2012, p 18-19
Adaptive monitoring of vegetation in the Macquarie Marshes, New South Wales
Office of Environment and Heritage, Dubbo, NSW. Email: Darren.email@example.com
Exclosure plot design showing the four treatments. Bold line = fenced. Dotted line = unfenced.
Dashed line = border between raked and unraked treatment. Unraked areas had an abundant groundcover of Sclerolaena spp. In raked areas the Sclerolaena was removed by raking.
Vegetation changes in unraked treatment outside Exclosure 3. (a) December 2008; (b) November 2009;
(c) March 2010; (d) November 2010.
Photos: Darren Shelly.
In 2005 a group of Macquarie Valley landholders formed the Macquarie Marshes Environmental Trust to purchase the 259 ha property ‘Burrima’ on the western side of the northern Macquarie Marshes. The property contains three distinct vegetation communities, each covering about a third of its area. The communities (from west to east) are:
- low chenopod (Atriplex/Sclerolaena) shrubland on cracking brown clays;
- Coolibah (Eucalyptus coolabah) open woodland with often dense Black Roly-poly (Sclerolaena muricata) understorey on cracking grey clays;
- River Red Gum (Eucalyptus camaldulensis) woodland on cracking grey clays.
After purchase the property was de-stocked, although it still contained some native and feral animals including kangaroos, emus and feral pigs. Its management regime became similar to that of the adjacent northern Macquarie Marshes Nature Reserve. The response of the vegetation to de-stocking has been monitored in an adaptive way, as outlined below.
Adaptive monitoring after de-stocking
Initial monitoring plots
Two 20 m x 50 m survey plots were established in each of the three vegetation communities on the property, to determine whether the removal of domestic stock leads to:
- an increase in overall floristic diversity, groundcover vegetation and perennial species diversity;
- a decrease in exotic species diversity and cover, and annual species cover/abundance.
In July 2005 (just after de-stocking) full floristic descriptions were obtained for each plot. The cover and abundance of each species was then recorded at six monthly intervals until January 2009, and thereafter recorded each January. The plots were affected by a major flood in 2011.
Grazing exclosure plots
Early in the monitoring program it was noted that grass re-establishment was poor and most tussocks were grazed close to ground level, presumably by native and/or feral animals. To monitor the impact of this grazing and investigate additional methods of groundcover establishment, in August 2008 one vertebrate grazing exclosure plot was established in each of the three vegetation communities, each plot with the same grass establishment treatments.
Raking to remove Sclerolaena was carried out in the ‘Raked’ treatment area before fencing. Each grazing exclosure (fenced area) was 72 m x 72 m with a 1.4 m high Weston design fence constructed in a way to prevent animals digging underneath or climbing through it.
Grass seeds were hand sown throughout each exclosure and in the adjacent unfenced areas, using seeds of Queensland Bluegrass (Dichanthium sericeum), Curly Mitchell Grass (Astrebla lappacea) and Warrego Summer Grass (Paspalidium jubiflorum).
Five randomly located 1 m2 plots were measured in each of the four treatment areas in each community. All vegetation was identified in each plot, the number and percentage cover of each species recorded, and the percentage cover of litter, bare earth and cryptogams recorded. Grazing levels outside the exclosures were monitored indirectly by faecal pellet counts. An initial survey was conducted in December 2008 to obtain a dataset as close as possible to the time of fence construction, then monitoring was repeated annually during spring.
Monitoring in the grazing exclosure plots and adjacent treatment areas aimed to:
- determine the impact of total vertebrate grazing pressure on natural grass establishment and the establishment of sown native grass species;
- investigate the impact of protection from vertebrate grazing on the cover and/or abundance of other groundcover species;
- investigate whether initial removal of the most abundant groundcover would aid native grass establishment.
Results to date
Initial monitoring plots (July 2005–January 2011)
The total number of species recorded from all transects increased from 57 to 72 over the six year period. Coolibah woodland showed an increase from 31 to 34 species, while River Red Gum woodland showed a decrease from 34 to 31 species, mainly due to its transects being flooded and aquatic plant richness being lower than for terrestrial species.
The low chenopod shrubland species richness hardly changed during 2005–08, then reduced significantly in drought but recovered after good seasonal rains to record its highest richness on the last monitoring period in January 2011. Over the entire period its species richness increased from 21 to 37 species.
Vegetative groundcover decreased in all communities, but there was a corresponding increase in leaf litter accumulation (in the woodlands) and cryptogam cover (in the low chenopod shrubland). The total groundcover thus remained at similar levels over the survey period.
Perennial species diversity increased in two of the three vegetation communities, the greatest increase being in the low chenopod shrubland (which experienced the least flooding). The 2011 floods decreased the number of perennials in the Coolibah and River Red Gum plots, although monitoring to 2010 showed steady increases in perennials over the 2005 levels.
Only the River Red Gum woodland recorded a decrease in the number of exotic species over the survey period. However, the variation in exotic species diversity was only in the order of one species for each community. All three communities showed some annual species decreased in cover and/or abundance over the six year monitoring period.
Grazing exclusion plots (December 2008 to September 2011)
Grass species richness, cover and abundance were higher inside the grazing exclosures than outside. Grazing pressure outside the exclosures was so high the effect of raking on subsequent grass establishment could not be assessed. After 15 months and a high summer rainfall, all three sown grass species grew and seeded within the exclosures; none was recorded outside.
Raking reduced the cover of the herbaceous chenopods (Sclerolaena spp.) for 12–15 months only, after which it returned to around pre-raking levels in response to good summer rainfall.
Over the monitoring period there was a gradual decrease in the proportion of bare earth both within and outside the exclosures. On the scalded brown clay soils, litter and cryptogams tended to replace bare earth rather than vegetation, although vegetation cover increased on the grey clay soils. It seems that kangaroo grazing per se does not significantly affect changes in groundcover characteristics over time as similar trends occurred both within and outside the exclosures.
The monitoring data have shown that after the removal of domestic stock, grazing by native animals such as kangaroos significantly reduces the recovery of native grasses and thus the replenishment of their soil seed banks.
With recent funding from the Central West Catchment Management Authority all vertebrate grazing animals (native and feral) will be excluded from nearly half of ‘Burrima’, allowing grasses and forbs to grow undisturbed for several years. Once the grasses are fully established the fencing will be removed to again allow kangaroo access. Their ongoing impact on the viability of grass species will be monitored. The results will have rehabilitation implications for marshland vegetation which typically retains high densities of kangaroos year round.
Thanks to Chris Hogendyk, Chairman Macquarie Marshes Environmental Trust, and all the committee members for allowing unrestricted access for monitoring and for encouraging all types of research and educational opportunities within the property.