Doctor of Philosophy
Department of Biological Sciences - Faculty of Science
Gosper, Carl R, Consequences of weed invasion and control on plant-bird interactions and bird communities, PhD thesis, Department of Biological Sciences, University of Wollongong, 2004. http://ro.uow.edu.au/theses/371
Introduced plants can have a variety of impacts in ecosystems in which they become invasive. These impacts can include the disruption of interactions between native plant and animal species, such as seed dispersal and pollination. Subsequently, other interactions and mutualisms can be affected, both at the site of the plant invasion and elsewhere. Interactions can also become established between the invasive plant and native and invasive animal species. The removal of an invasive plant has the potential to disrupt these newly formed interactions, thus disadvantaging some fauna and potentially affecting subsequent interactions involving these species. While control of invasive plants is typically a goal of conservation management, the consequences of control for other species are rarely fully considered or investigated. In this study, I have tested several hypotheses regarding the interaction of an invasive plant with fauna, and the effects of the plant’s invasion and control on plant-animal interactions. Bitou Bush (Chrysanthemoides monilifera ssp. rotundata) is an invasive plant in coastal New South Wales (NSW), Australia. South African in origin, it has now spread to occupy at least 80% of the NSW coastline. Fruit production in C. monilifera is prolific and fruits are consumed and dispersed by vertebrates, especially birds. In addition to other effects of C. monilifera, this plant-animal interaction has the potential to affect the seed dispersal of other vertebratedispersed plants and also the composition of the bird community, which may subsequently affect other plant-bird interactions. In order to quantify the magnitude of these possible effects, I designed this study with the following five major components: (i) comparison of the fruiting characteristics of C. monilifera with those of co-occurring birddispersed native plant species; (ii) description of the plant-bird interactions that involve flowers and fruits in vegetation that has been invaded or is at risk of invasion by C. monilifera; (iii) measurement of the rate of removal of C. monilifera fruits and those of some co-occurring bird-dispersed plant species in: habitat dominated by C. monilifera, where C. monilifera had been eliminated by the application of herbicide, and uninvaded vegetation; (iv) assessment of the effect of C. monilifera removal by herbicide application on the species composition and abundance of the bird community; and (v) assessment of the effect of dominance of the vegetation community by C. monilifera on the species composition and abundance of the bird community. I found that the fruits of C. monilifera are within the range of physical dimensions and nutrient composition of those of co-occurring native species. The greatest distinction in fruit characteristics is in phenology and the combination of phenology and morphology, as peak production of C. monilifera fruits occurs when native fruits are scarce. Consequently, C. monilifera fruits are attractive to vertebrate dispersers, especially birds. At least 25 species of birds feed on C. monilifera fruits in NSW, most of which are indigenous and are likely to disperse viable seeds. In an experimental study using feeding stations, I found that the rate of removal of fruits of native plant species was unaffected by either dense infestation of C. monilifera, or its elimination. This is likely to be due to highly facultative relationships between frugivorous birds and plants, combined with differences in phenology and, in some cases, the morphology of fruits of native plants and C. monilifera. Consequently, in this system there has been little impact of an invasive plant on this plant-bird interaction. The rate of removal of C. monilifera fruits, however, was less in herbicide-treated habitat. This has implications for long-term C. monilifera control, because herbicide treatment often leaves scattered individual plants alive, which would have poorer seed dispersal than plants in dense C. monilifera stands. The removal of C. monilifera affected the bird community, with the overall abundance of birds declining in herbicide-treated areas after the C. monilifera died. Only those birds that consume C. monilifera fruits were affected, while other groups of birds that do not directly use C. monilifera resources were unaffected. Although this impact was minor, it illustrates that removal of an invasive plant can affect bird communities, and these impacts should be considered before control programs are implemented. Dominance of the vegetation by C. monilifera also affected bird communities, with overall bird abundance, specifically that of insectivorous birds, and at some locations nectarivorous birds, being lower in C. monilifera than native habitat. The quantity of remaining native vegetation, particularly of nectar-producing plants widely used by birds, appears to be important in determining bird community composition in invaded areas. In this study, I have demonstrated that both dominance of the vegetation by an invasive plant, and the control of an invasive plant, can induce change in fauna communities, and disrupt some plant-animal interactions. These changes will need to be considered carefully in planning management actions to conserve coastal bird communities and their interactions with plants. While continued efforts to control C. monilifera are clearly justified, these should form part of a broad strategy for coastal community conservation, including consideration of other threats to native communities that act independently or in concert with C. monilifera invasion. These considerations should include the potential impacts of other invasive plant species, targeted sites and species for control efforts, and other forms of habitat loss and degradation.
Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.