Once the dominant vegetation cover in what is now western Sydney, Cumberland Plain Woodland is a recognised endangered ecological community which exists in small fragments within and between Sydneys western suburbs. It is characterised by the coexistence of Themeda triandra, Eucalyptus tereticornis, Eucalyptus moluccana, Brunoniella australis, Bursaria spinosa, Glycine tabacina and Dichondra repens. Second only to habitat loss, weed invasion is the largest threat to this community. Olive (Olea europea ssp. africana) and broad-leafed privet (Ligustrum lucidum) are two woody weeds that are commonly found in Cumberland Plain Woodland fragments. They are spread by fruit-eating birds which perch in branches and drop seeds in their faeces. This results in a dense infestation of olive and privet around a central perch tree, often a eucalypt. It is of great concern that the invasion of these weeds threatens the biodiversity of an ecological community that hosts several endangered ROTAP species including Pimelea spicata and Acacia pubescens. In order to preserve this ecological community, regeneration of the local native species is required following control of olive and privet. This project aimed to investigate some of the impacts of the accumulation of olive and privet leaf litter and to explore how it may affect the germination and early growth of native species. In order determine the impact of weed litter, the litter of two native species, namely kangaroo grass (Themeda triandra) and bursaria (Bursaria spinosa) were included in each of the experiments as a comparison. The work was divided into two components; field sampling and a glasshouse experiment. These were designed to explore the physical and chemical effects of the litter.
For each of kangaroo grass, bursaria, olive and privet, 4-6 sites were chosen in fragments of Cumberland Plain Woodland. Where possible sites were selected where more than one litter type occurred. Within each site three quadrats of 0.25 x 0.25 m were laid out and the depth (cm) of the leaf litter measured. Litter within the sample quadrats was then bagged and later dried and weighed (g). Following the removal of the litter, two soil cores up to 10cm in depth were taken. Within 1.5m of each quadrat a pit fall trap was set up. Traps consisted of plastic cups that were buried so that their rims lay flush with the soil surface and they were partially filled with a detergent solution. The traps were set out for three days, after which their contents were drained and stored in 70% ethanol. During the selection of sites it was determined that olive is more abundant and widespread than privet in fragments of Cumberland Plain Woodland.
The litter depth and mass of the weed species was significantly (P £ 0.05) greater than that of the native species. This may affect seed germination and may be associated with more intense fires in olive and privet infestations during controlled burns or bushfires (A.M. Gill, pers. comm)
Soil samples were dried, sifted and analysed for pH, inorganic nitrogen and available phosphorus. Soil from beneath weed litter with an average pH of 5.8 was significantly (P £ 0.05) less acidic than that from beneath kangaroo grass and bursaria, which averaged pH 5.1. There was a large amount of variation in the inorganic nitrogen results, with no significant difference between the soil sampled beneath any of the litter types. Privet samples were associated with significantly (P < 0.01) higher levels of available phosphorus than any of the other litter types. In fact, phosphorous availability beneath privet infestations was about 5 times higher than its availability in soils dominated by native vegetation. This result requires further confirmation and investigation.
The invertebrates caught in the pitfall traps were primarily sorted into ants, spiders and other. Due to the large number of ants present in each sample, these were selected as invertebrate indicators of environmental change that may be associated with the invasion of olive and privet. Having divided the ants into morphospecies, diversity was found to be greater by 2-3 morphospecies in native litter than in olive and privet. The spiders and other invertebrates are yet to be sorted.
A three way factorial experiment was set up to investigate the effect of leaf litter from different plant species on the germination of a selection of native plants. The experiment was designed to expose any effects that litter may have on germination and to determine whether the effect, if any, was chemical or physical. The first experimental factor was seed species. Themeda triandra, Eucalyptus tereticornis, Bursaria spinosa, each a common Cumberland Plain Woodland species, and Pimelea spicata and Acacia pubescens, both endangered species, also from this ecological community, were used. The second factor was litter type. The four litter types sampled in the field and described above were used; namely, litter from habitats dominated by kangaroo grass, bursaria, olive and privet. The final factor was litter treatment where leached and unleached litter was used. Leached litter was achieved by soaking the litter in water for three days, to remove any chemicals that may affect germination. This treatment would allow distinguishing of chemical and physical effects. A treatment using no litter was included for comparative purposes.
Black plastic pots of 10 cm diameter were used. Once filled with soil, twenty seeds were scattered on the surface which were subsequently covered with the required litter to a depth of 2cm. Six replicates of each treatment combination were set up in a fully factorial, randomised block design. Germinations were scored twice weekly, and all germinants were removed with the exception of the first in each pot which was left to grow, thereby enabling any effect on early growth to be determined at a later date
There was no significant difference in the percentage germinations of any of the five seed species between any of the litter types including the leached and unleached litter. The seeds that had no litter covering them, however, showed significantly fewer germinations.
As mentioned one seedling was left in each pot. Their growth will be continued for approximately eight additional weeks, after which they will he harvested and their dry mass determined.
The results of the project suggest that olive and privet invasion is definitely a threat to Cumberland Plain Woodland communities. The leaf litter from these species is deeper and has a larger mass which has implications for burning as a possible weed control tool. Litter of the studied woody weeds significantly increases the soil pH and available phosphorus. The changes in pH is important as it can affect the availability and abundance of many other chemicals and nutrients. While privet appears to be less abundant than olive, it may be the most harmful due to the higher pH and available phosphorus in the soil beneath its litter. The preliminary sorting of the invertebrates suggests that the invasion of these woody weeds may affect both faunal and floral biodiversity. Once the remaining invertebrates have been sorted this result may be strengthened. It was positive to determine that the germination of the native species selected were not inhibited either physically or chemically by the weed litter. This project has provided a good initial investigation into some of the effects of olive and privet invasion on Cumberland Plain Woodland species which may impair regeneration. It has provided a general overview of several aspects and highlighted some which are likely to be more important, such as soil phosphorus.