To observe changes over time we set up permanently marked plots that we could periodically record. Initially we set up eight permanent square plots, each 5 m x 5 m, marked with steel posts and coloured tape. The Woodland Conservation Area included an area with large trees which appeared likely to have a good number of native species amongst the weeds, as well as areas with smaller trees and more open areas that had evidently been cultivated or possibly ploughed and pasture improved. When we first visited it had been grazed by stock for many years and had been recently slashed.
Historical accounts give the impression that Themeda triandra, Kangaroo Grass, was a major groundcover component of the woodland at the time of European settlement, but that this was readily eaten out by stock. Patches of Themeda triandra were relatively common in our woodland and were relatively weed free. We regarded these as the type of groundcover we hoped would expand into the more weedy areas in the absence of grazing and mowing. We placed two sites in grassland mainly of Themeda triandra (sites 3 & 4), and two sites in Themeda triandra-Bursaria spinosa, Blackthorn, grassy-shrubby areas (sites 1 & 6). To sample the weedy areas we placed two sites in *Paspalum dilatatum-*Sida rhombifolia, Paddy's lucerne, dominated areas (sites 2 & 5). Because of our particular interest in the endangered species Pimelea spicata we located one site to include part of its population, (site 7), and another, (site 8), amongst Chloris ventricosa-*Sida rhombifolia-Bursaria spinosa, with a similar shrubby structure but with no Pimelea spicata.
Initial field recording was done in December 1988 as we thought that most plants would be flowering late spring-early summer and most readily identified. We recorded the species present within each of the 25, 1 m x1 m subplots. This gave us a frequency out of 25 for each plot, and a total frequency out of 200 subplots, a measure of abundance that we could remeasure, to observe increases or decreases. Our 8 plots set up in 1988 included occurrences of about 50 % of the plant species (69 out of the 134) recorded in the woodland at that time.
Some difficulties often encountered in monitoring Cumberland Plain Woodland
- Dry conditions seasonally and even daily.
- Tiny species make location and identification difficult, particularly if they are also short lived e.g. Phyllanthus virgatus.
- Similar closely-related species make identification difficult e.g. Austrodanthonia species.
By 1991 rabbits were becoming common in the woodland and despite the assurances of the Gardens' management that a control program would eliminate them within 6 months, evidence of their activities became increasingly obvious. We had not considered the impact of rabbits during the initial setting up of the plots as there was no evidence of their presence, and the open conditions of the grazed landscape and presence of dogs presumably kept them in check. However the taller growing groundcover and the fencing of the Gardens obviously provided them with suitable conditions to flourish. In 1991 we set up three fenced exclosure plots, Plot 9 paired with Plot 6, Plot 10 paired with Plot 8, and Plot 11 in open grassy area at the top of the woodland, paired with a new unfenced plot 12. We particularly wanted to see whether natural recruitment of native seedlings was retarded by rabbit grazing. These plots were recorded at the same time as Plots 1-8.
Subsequent recording of all the plots was done on an ad hoc basis as time allowed. The plots were recorded in May 1991, November 1991, December 1992, December 1993 and December 1996.
After the 1996 recording we began to question our expectations of the change in the woodland. There was not the great expansion of native species we had expected. Instead change was very slow. What was beginning to happen however was that some of the perennial weed species were beginning to expand. This was obvious in observations in the woodland, and was confirmed in the plots. In particular the weedy shrub *Olea europea subsp. cuspidata, African Olive, was clearly becoming more common in the woodland, particularly under large Eucalyptus trees. The time spent on the sample plots seemed disproportionate to the questions being raised, which related to rare species, that were mostly not represented in the plots. Other work commitments reduced our available time for the project, allowing only occasional visits to the woodland and not allowing for the week needed every year to record all 12 plots plus the time to consider the data. No recording was made from 1997 to 2000 though we visited the woodland sporadically during this time.
Following changes in work programs at the end of 2000, we were able to allocate more time to the woodland allowing us to record our sites in January 2001, December 2001, December 2002, December 2003, December 2004 and December 2005.
In 2000, after reviewing the woodland plot data from the previous 10 years, during which relatively little seemed to have happened, we became concerned that our narrow window onto the woodland, our December recording, meant that we were missing important changes in the vegetation that occurred in other months. We had initially chosen December for annual recording to maximise our level of identification, based on our wider experience on sandstone areas where major flowering is in spring, and modified by knowledge that grasses are more readily identified by their fruiting in summer. We realized that many of the plant species were opportunists and that major ecological events such as seedling recruitment, flowering and fruiting, were being influenced by rainfall and temperature events which occurred irregularly over the year.
We therefore selected two of our plots, Plots 4 and 9, for regular monthly recording. Plot 4 was an unfenced grassy plot which was left undisturbed without any treatment. Plot 9 was a fenced Themeda-Bursaria plot set up in 1991. Neither plot has been burnt in the 20 years to 2006. These two plots contained a good variety of woodland species, 67 natives, (50% of our total of 134) and 34 exotics.
Our more frequent recording subsequently has shown that overall there is more growth and flowering after rains in late summer and autumn than there is in spring, although a handful of species grow more in spring.
To explore the issue of seasonality further we have been recording the flowering and fruiting times or seasons of the woodland plants. This is known as phenology. Our initial view that flowering and fruiting timing were likely to be similar to the patterns observed on the familiar Hawkesbury sandstone areas was changing, and we wanted to observe changes in our woodland more closely.
We carried out a trial ecological burn in part of the woodland in 1991.
We carried out more experimental burns in parts of the woodland in September 2001 when we again burned the area previously burnt in 1991. The adjacent area to the west was burned in September 2002, the area to the north, including adjacent grassland, in September 2005.
Asterisk * indicates exotic species naturalised at the Australian Botanic Garden.