The effect of soil disturbance, grazing and ground cover compositon on the expansion of an invasive plant, broom Cytisus scoparius, along the Shoalhaven River in the Southern Tablelands of New South Wales, Australia

  • Published source details Sheppard A.W., , P. Paynter Q. & Rees M. (2002) Factors affecting invasion and persistence of broom Cytisus scoparius in Australia. Journal of Applied Ecology, 39, 721-734


Broom Cytisus scoparius, a native European leguminous shrub, has been introduced to several areas of the world. In parts of its introduced range in Australasia, and North and South America, it has become of environmental concern due to its invasiveness.

In south-eastern Australia broom is invasive in habitats including upland grasslands, open woodlands and along water courses where it suppresses native plants and reduces the potential of agricultural grazing land. The basic ecology of broom at the time of this study, with the exception of recruitment, was fairly well understood. An experiment was undertaken to understand the dynamics of broom recruitment in several areas of grassland. Its aims were to evaluate the roles of disturbance, grazing and ground cover composition on the rate of broom colonisation, and regeneration within mature stands following removal.

Study sites: Two localities, 50 km apart, along sections of the Shoalhaven River system infested with broom Cytisus scoparius were selected:

1. Krawarree (600 m a.s.l.) (2 sites) – At this locality broom had formed a 5 ha solid stand across an unfenced border of cattle pasture into the Deua National Park. The farmer ceased broom management 14 years prior to the experiment. A fence was erected to prevent cattle entering the site 2 years before the start of the experiment. It was accessible to numerous kangaroos and wallabys (Macropodidae), wombats Vombatus ursinus, feral pigs Sus scrofa and rabbits Oryctolagus cuniculus.

Site 1 (Kwrawaree improved) – two pairs of plots were established, one in mature and one in immature broom, both on the improved pasture side. The pasture was dominated by another non-native, the perennial grass Phalaris aquatica. One plot of each pair was fenced (mesh size 5 cm, 1.5 m high, buried to 50 cm depth) to exclude cattle and other mammalian herbivores.

Site 2 (Kwrawaree native) – two unfenced plots in mature and immature broom on the national park side of the broom stand, where the native grassland is dominated by Poa labillardieri, Microlaena stipoides and Themeda australis.

2. Waterboard (500 m a.s.l.) (Site 3) – A cattle- and sheep-grazed paddock on the west bank of the river where broom of mixed age had formed a 30 m wide, dense strip along the river, separated by it from a 30 m strip of blackberry Rubus discolor. The grassland was dominated by natives mainly Poa labillardieri, Carex gaudichaudiana and Microlaena stipoides and exotic annual grasses.

Three pairs of plots were established (one each being fenced as at Kwrawaree) in mature broom, immature broom and 5 m out from the edge of the broom strip on the pasture side.

Treatments: Removal and disturbance treatments were randomly assigned to three 5 x 5 m subplots within each of the plots.

In two subplots all broom growing in or over-hanging the subplot was cut to ground level. In one of these, stumps were painted with Grazon® (active ingredients Triclopyr and Picloram) leaving the herb layer undisturbed, as happens after natural senescence (cut treatment). In the other, all stumps were removed and the ground manually cultivated to 10 cm depth, resembling disturbance caused by pigs and wombats (disturbance treatment). The third subplot was left untouched as a control.

This resulted in one subplot each in mature and immature broom per grazing x disturbance treatment combination at each site except at site 2 where there was no ungrazed treatment, and at site 3 where an additional 5 m from stand edge, maturity category was used.

Seed sampling: In November 1993, before treatment, 10 soil cores (32 mm diamater, 10 cm deep) were taken from each subplot. These were sieved, sorted and the number of seeds in each was recorded. A glasshouse germination experiment was then conducted, the seeds being combined and placed on moist filter paper, watered regularly for four weeks and recording the number that germinated. Those exhibiting no signs of germination were cut to detect fresh endosperm. The seed bank could then be divided into germinable, viable and dead categories.

Soil cores were subsequently taken twice a year. The first sample was taken just before seed fall in the summer to give an estimate of the minimum seed bank. The second was taken after seed fall and prior to autumn germination to give an estimate of annual seed fall.

Seedling sampling: In each subplot five permanent 0.5 x 0.5 m quadrats were randomly positioned but if none fell within areas of high seedling density two additional quadrats were selected. Seedlings were counted and position of the 1-year old plants recorded.

Pod sampling: The number of pods per plant were counted in the summer. All or a sub-sample of 30 pods per plant (for those plants with greater than 30 pods) were collected and seeds per pod counted.

Ground cover estimates: The percentage cover of all plant species, litter and bare ground was recorded in five randomly positioned 0.5 x 0.5 m quadrats. Theses were re-randomized on each sampling date.

Sampling dates: Seedling and pod sampling, and ground cover estimates were undertaken in spring (September-November), autumn (March-June) and after each major rainfall (> 10 mm) in summer (November-March). Sampling was not undertaken during droughts in summer 1994-95, autumn 1996 and the El Niño drought from July 1997-May 1998. Sampling commenced in 1994 and continued until March 2000.

Additional observations: During sampling, presence and damage by arthropod herbivores, molluscs and fungal pathogens was noted. Over the whole seven-year sampling period negligible impact levels were recorded.

Initial seed banks: In the mature broom plots, the seed bank ranged from 4,229 to 21,012 seeds/m² across the three sites. At site 2 in native grassland adjoining Deua National Park, Krawarree, the seed bank was about one quarter that of the other two stands. The seed banks increased on average from 1,425 to 9,117 seeds/m² in the spring and autumn samples equivalent to an increase of 17-68%.

Seed banks were about three to 10-fold lower at 1,419-2,396 seeds/m² in the corresponding immature broom plots, and 5 m beyond the immature broom only 39 seeds/m².

Germination experiment: 15-21% of seeds germinated within the 4-week duration of the glasshouse germination experiment. Seed bank viability was similar across all sites at 96-98%.

Subsequent seed banks according to treatment: Seed banks declined significantly in the cut/disturbed mature broom subplots during the first 4 years after treatment to 14% of their original densities. After this regenerating broom began to flower and produce seed, the seed bank decline ceased. In the immature broom plots seed banks at some e.g. site 3, appeared to increase over the six years of the study, presumably as plants matured and more flowers and seed were produced, but overall patterns were not significant.

Seedling germination & survival: There was little seasonality in rainfall at any of the sites and fresh seedlings were seen in some subplots in each sampling period. It is considered that the frequency of censuses (three times a year) was insufficient, missing much germination which probably occurred throughout the wetter months. The proportion of seedlings surviving 12 months varied significantly between sites (Krawaree improved - 0.27; Krawaree native – 0.05; Waterboard – 0.05) and either disturbance type more than doubled survival compared with the undisturbed controls. The effect of grazing was similar. The higher survival rate at Krawaree improved may be explained by the broom cover and more fertile soil enhancing survival in the short term.

Drought was considered to be the main cause of seedling mortality with greatest losses in shallower or river-sand –based soils i.e. where water retention properties were lowest. Survival over 24 and 36 months was also heavily influenced by such drought conditions.

Flowering: Age of broom plants at first flowering ranged from 2 to 5 years with 1.3% flowering at 2 years, 10.9% at 3 years, 80% at 4 years and 8% at 5 years old (n = 386). The percentage of plants producing viable seed increased with age at first flowering: 20% at 2 years, 50% at 3 years, 73% at 4 years and 87% at 5 years old. The number of seeds per pod was not affected by any factor and averaged 7.75/pod. Fecundity (number of pods x number of seeds/pod) was unaffected by site, stand maturity, grazing, disturbance or cohort, but was related to plant density in first and second years of flowering. This density dependent flowering was stronger in the second year. The average seed fall between cut and cultivation treatments was:

1997-1998 (4th year) - 8 and 11 seeds/m²
1998-1999 (5th year) - 148 and 243 seeds/m²
1999-2000 (6th year) - 329 and 769 seeds/m²

Percentage cover: There was no significant difference between mature or immature broom plots of percentage broom cover, competing vegetation, litter and bare soil in the subplots across sites prior to treatment allocation. Perennial grasses and sedges dominated all sites. Following broom removal, whether the ground was cultivated or not had little effect on the speed of regeneration. This indicated weak suppression from the ground flora in the uncultivated plots. Total cover of competing vegetation was unaffected by cut, cultivation or grazing treatment, although cultivation reduced the perennial grass cover and increased cover of annual and perennial herbs. Grazing showed a tendency to increase the rate of broom regeneration.

Overall, broom cover was low for the first two years in cut or cultivation plots but by the sixth year cover had returned to initial levels. In grass-dominated plots however, regeneration took more than six years.

Conclusions: Overall the evidence from this study suggests that the presence of competing vegetation has little effect on broom regeneration. However, immature plant mortality can be high, particularly in native grasslands and also during times of drought. It appears that successful control of broom in Australia is likely to be site dependent with no one overall management strategy that can be prescribed for eradication.

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

Output references

What Works in Conservation

What Works in Conservation provides expert assessments of the effectiveness of actions, based on summarised evidence, in synopses. Subjects covered so far include amphibians, birds, terrestrial mammals, forests, peatland and control of freshwater invasive species. More are in progress.

More about What Works in Conservation

Download free PDF or purchase
The Conservation Evidence Journal

The Conservation Evidence Journal

An online, free to publish in, open-access journal publishing results from research and projects that test the effectiveness of conservation actions.

Read the latest volume: Volume 17

Go to the CE Journal

Subscribe to our newsletter

Please add your details if you are interested in receiving updates from the Conservation Evidence team about new papers, synopses and opportunities.

Who uses Conservation Evidence?

Meet some of the evidence champions

Endangered Landscape Programme Red List Champion - Arc Kent Wildlife Trust The Rufford Foundation Save the Frogs - Ghana Bern wood Supporting Conservation Leaders National Biodiversity Network Sustainability Dashboard Frog Life The international journey of Conservation - Oryx British trust for ornithology Cool Farm Alliance UNEP AWFA Butterfly Conservation People trust for endangered species Vincet Wildlife Trust